Hi, I'm an electrical engineer, mostly high power DC/DC converter stuff where thermal engineering is an absolutely vital part of the design. So I like to think I have at least a little expertise when it comes to thermal management and cooling.
First off, everyone needs to post their ambient room temperatures at the time they measured the CPU/etc. temps. Without knowing the ambient temperature of the room at the time you took those measurements, we can't really be sure about the improvement in cooling performance.
For those interested, I have provided a short little primer on basic heat for those interested. If you don't really care, just skip this section and read the bold TLDR 
.
Please, the first and most important temperature you should post is that of the room the computer is in, and for each different time you measured, say, CPU temperature.
@samanosuke 's screen caps, for example, could be the exact same cooler when the room is slightly warmer. The slightly cooler measurements might actually represent inferior cooling to the hotter one, if the room was a few degrees hotter during the first measurement. Most of us probably have some sort of temperature control, but even then, you'd be surprised at how much the temperature can fluctuate in a room throughout the day.
Heat is one of those things that few people have a good understanding of, and understandably so. You can't really see it, it moves in vague and unintuitive ways, and there is no metaphor to describe it, like using water for electricity. It's easy to get the gist of heat, but its very hard to really understand heat (personally, it took me years to develop an intuitive understanding of it). So, I will do my best to explain, but don't worry if you don't follow. But please take my word for it
.
Wattage or TDP (total power dissipation) causes temperature RISE. Given any fixed wattage being dissipated by a given object, that object will eventually reach equilibrium (it is releasing heat as fast as it is generating it) and it will stop getting hotter. The temperature will have RISEN <X> number of degrees above room temperature. This is called thermal resistance, and it is exactly what it sounds like. It is the resistance to the flow of heat. It doesn't have a unit, but is usually given in degrees per watt, which means that, say, if a chip has a thermal resistance of 2 degrees per watt, that means its temperature will increase 2 degrees above room temperature for every watt of heat it is producing. If it is producing 20W, it will heat up 40 degrees.
Nowhere in any of that have I one mentioned what temperature it will actually reach, because we don't know. We only know how much the temperature will INCREASE. We have to know the room/ambient temperature to be able to tell what the thermal resistance is. And thermal resistance is ultimately the thing that matters. Given two cooling setups, the one that lowers the thermal resistance more is the better one, and that is also how you can tell how MUCH better.
TLDR; The actual temperatures of your CPU do not indicate cooling performance. Temperature difference does. To actually measure performance of a cooling setup, you must report how many degrees hotter the CPU is than the room your computer is in, and at the specific time you measured the CPU temp. The actual temperature of the CPU alone is not useful.
This brings me to my main point, which is that liquid cooling is, at least for computer cooling applications, slightly inferior in most performance metrics compared to air cooling. A properly designed case with adequate fan layout and number will always be slightly cooler, quieter (yes, quieter!) and reliable than liquid cooling.
Liquid cooling's strength is that it removes most of the potential for human error as well as poor case design, as factors in cooling. But remember, liquid cooling
IS air cooling. There is still a fan, and a heatsink-like radiator. Liquid cooling is nothing more than the way you are moving heat from the components inside the case to something that exhausts heat, via a fan, out of the computer. Liquid cooling concentrates the heat in the radiator which pulls in air from the cause, blows it over the hot radiator, exhausting hot air out of the case.
Because it uses liquid to move the heat, you just have to connect the tubes and pump the right way and it works, it is relatively foolproof. As long as the liquid is moved from point A to point B, it works.
Air cooling is the same, but the heat is moved by air, and in a less controlled way, so case design and user error in fan placement/direction can significantly impact air cooling where liquid cooling cannot be so impacted.
However, if you know what you are doing and selected a good case (easier said than done), you can always get better results with regular air cooling.
Air cooling is entirely dependent on CFM, or the shear volume of air that is sucked into the case, hopefully blown over the heat generating sections, and then exhausted out of the case. The closer you can make the inside of the case like a giant wind tunnel, the better. It should exhaust all in one area, and as far away as possible from the fans that pull in air. Otherwise, you will be sucking in the hot air you just blew out.
You might think that a liquid cooling block would be infinitely more effective than the copper heatsinks used in air cooling, but that is not the case. Those heatsinks and the silver or copper colored rods/tubes you see all running towards the CPU are amazing devices called
heatpipes. Heat pipes are basically magic.
Heat pipes have the highest thermal conductivity of any known material. By 100 fold. The thermal conductivity of diamond, which is the highest known and roughly 3 times that of copper, is about 1,000 W/m•K.
Heat pipe thermal conductivity is anywhere between 5,000 to 200,000 W/m•k. In fact, larger industrial heat pipes can achieve thermal conductivities as high as
4,000,000 W/m•K.
Like I said, magic. No bulk material even comes close, and neither physically moving water being pumped through cooling blocks.
Those tiny little copper tubes are actually filled with a liquid (usually methanol for computer applications, but different temperature ranges use different fluids, including plain old water). But, heat pipes do not move heat via liquid. At the hot end, the fluid is heated above it's boiling point, turning it into a gas. This absorbs a tremendous amount of energy. Imagine how much heat you can dump into a pot of boiling water? No matter how high you turn the stove, it won't get above boiling, but might just boil a little harder. Water merely being pumped through tubes is nothing compared to the heat absorption achieved via the liquid's
Enthalpy_of_vaporization (latent heat). As the fluid expands to a gas, it will move to the cooler end of the tube, carrying all that heat with it, where it condenses back down to a liquid. The insides of a heat pipe are sintered copper powder, which make it sort of like a copper 'sponge', and this will
wick the liquid back towards the hot end, like water soaking into a sponge when just part of it is submerged. This entire cycle is continuous, and that means that those CPU heatsinks with fans are, in reality,
phase change coolers that move heat by actually changing liquid to a gas, and back again. Phase change cooling is vastly superior to any form of liquid cooling, do to shear laws of physics.
So, a proper heatsink with good heat pipes completely outclasses anything achievable with a liquid cooling block, at least in terms of removing heat from the CPU. It moves it into the hundreds of fins that make up the heatsink, providing ample surface area for air to then move that heat out of the heatsink. Which is no different than liquid cooling: it is using a tube or pipe or sorts to move heat to a heatsink/radiator that has its heat removed by a fan. Only, a CPU heatsink uses phase change cooling to move the heat, which absolutely destroys liquid cooling in every measurable way.
Air cooling seems inferior usually because its the part after this, where the heat needs to be moved out of the heatsink, that gets messed up. You want to suck air through the heatsink, not blow air through it. You want to suck air from the cool air being sucked in by intake fans, and blow it out towards exhaust fans that exhaust hot air. And you want to use many large but slow moving fans so your air flow is smooth like a brook instead of turbulent like white water rapids.
When properly done, your fans will be whisper quiet, and your temperatures will be cooler than what you can achieve with liquid cooling. And if you make sure to use air filters on your intake fans, and have more intake fans than exhaust fans (which will put your case at positive pressure, so all the holes that air can escape through will be exit only), then you will see virtually no dust build up over time. Dust can't get in if the only air that can get inside is air that is filtered.
Just to prove what I say, let me demonstrate:
I have a dual processor Xeon v4 machine, with 2 150W TDP Xeons, along with dual GPUs that produce an additional 165W of waste heat each (at full tilt). The i7-4790K has a stock TDP of 88W, and closer to 100W during all-core turbo, and the most heavily overclocked ones I've seen generally have a TDP of around 140W. So I think it is fair to compare one of my 150W Xeon CPUs to a highly overclocked i7-4790K.
The ambient temperature in my room is about 72° F. Fairly cool. Or, 22°C.
Under a full CPU and GPU stress test, both of my CPUs achieve indefinite all-core turbo, and I've never seen the die or a single core ever exceed 50°C. I am not sure about GPU temps. I specifically chose 'whisper quiet' editions and they are definitely whisper quiet, I am quite pleased. But they have their own cooling solutions built in so I'm mostly going to ignore them.
The CPUs on the other hand, I am cooling TWO 150W packages and never seeing temperatures above 50°C. This is a 28°C temperature rise. The ambient temperature inside the case is usually 25-28°C. This is because all of the heat is being quickly carried out so never has a chance to build up inside the case.
I am using midrange, single fan Noctua coolers, ones with relatively poor airflow, 78CFM I believe. I am also using 6 140mm fractal design fans, all of which have extremely crappy air flow individually, but they suck in from the bottom and blow out the top, creating an unobstructed wind tunnel, and the CPU and CPU coolers are aligned so they blow with, rather than against this airflow. The fans all have crappy CFM because I went for super quiet ones, all of them have a maximum noise level of 17dBA. For comparison, a whisper heard from 5 feet away is 20dBA. Now, that is the MAXIMUM noise level. I made sure to use a PWM controller (or, if your motherboard has the needed connectors, just use 4 pin PWM case fans, but I used 3 pin ones and a controller, which seems to be working wonderfully) and this will make them only blow as hard as needed.
And at full tilt, they all blow at about 600rpm. My computer is not silent, but it is far enough away that I can't actually tell if it is on or not by sound, I have to look over and check if the light is on or put my hand over the top and feel for air flow.
I am using an almost unreasonably large case though (I didn't quite realize how big it was when I ordered it...
) but it is very open and has terrific airflow. Specifically, its the Phantek Enthoo Pro.
TLDR Deux: Properly done air cooling will outperform liquid cooling every time, both in noise and temperature. However, it requires a case have enough open space and a design that can accommodate lots of 140mm slow moving fans, and all pointed in the correct direction and the internal CPU and GPU cooling fans aligned with this air flow. For more compact or dense computers, liquid cooling is almost always superior, but you can do even better with air cooling if you don't mind a slightly larger build and doing some careful planning and buying some fans. So keep that in mind when doing a build, and don't immediately discount pure air-cooling as inferior. It isn't. But it's easier to mess up, while liquid cooling will deliver consistent, relatively fool-proof results. Both are good in their own ways, both are trade-offs. 
(stock speed)! I know you can't really compare the two and maybe those temps weren't all that bad, but like i said having never owned K series processor it sparked me to try liquid cooling as i did not want to damage the new and a little more expensive CPU! I decided on the Corsair H100i GTX (with Thermaltake TG7)and just installed it tonight (which went very well quite easy to install) and idle temps dropped 11c and never got above 54c stock speed(haven't tried overclocking yet!) running GeekBench 3 64bit!!!!! Needless to say i am very impressed and satisfied with the liquid cooling and it looks great inside my modded G5 case!!! So like it says in the title list your CPU and what cooler you use with your temps would like to know where I stand temp wise!!!!
WHEN i do i'll post back work is super busy right now! Have had zero fan or pump issues though!!!
