Looks

This part is highly subjective so I took a couple of pics so you can decide for yourself. If you'd ask me, I like PCmod's more because the 4 knobs are centered (Maelstrom are a bit of to the left). PCmods changed the looks of their rheobus in V2: they changed the knobs (thank god, the ones supplied with V1 are really a bit too... dull), and changed the colours of the leds from blue to bright red. I'd rather have the blue leds, who wouldn’t?), but they are socketed and easy to change. Maelstrom doesn't have leds, which is not necessary of course. Bright blue leds are very pretty, but they can by annoying too when they light up your entire room while you’re trying to get some sleep. :) The knobs supplied with the Mealstrom look a bit unusual: some will like it, some will not. Personally I'd replace them right away.

Pc mods' (left) attached to 5 1/2" bay & the Maelstrom (right) naked. =)

Testing

Ah, the most important part! Most people don't know this, but the use of a rheobus will inevitably mean a loss of voltage. I don't know much about physics, but apparently there is no way to avoid this. However, there are different types of voltage regulators, some will do better than others. The Maelstrom uses a LM317 voltage regulator, which allows a theoretical maximum output of 11.25V (on a 12V line). I've seen these chips before in other kits, seems like it is a quite popular one. The PCmod's rheobus uses a TIP31 voltage regulator. I haven't found the max theoretical output of it, but I did find the spec sheets. Which one of the two is the best? Let's see how both handle a high performance 120mm fan (+- 6.5watts), a high performance papst NGH 80mm fan (2.8watts) and a delta medium performance 0812HH 80mm fan (3watts).

Test setup:
Abit KR7A-R
350Watt Enermax
RPM connected on mobo, ground and 12V separately on molex or rheobus.

120mm high performance fan

The multimeter I used is one of the cheapest analogue meters you can get, so don't expect the readings to be more accurate than 0.2 volt. First I connected the 120mm directly on a molex connector, the multimeter showed 11.50V (MBM: 12.19). Since the rheobus is connected on a molex, I don't expect the max output to be higher than 11.50V. :) Here are the test results in a nice little graph:

It seems like max voltage is greatly affected when using a rheobus, but my multimeter is pretty inaccurate so I did the same test again, but this time I checked the average rpm instead of voltage:

This graph endorses the measured voltage: max performance is greatly affected when using a rheobus. PCmods' TIP31 voltage regulator seems to do the job a little better than CaseEtc LM317, but still is far from perfect.

Of course, none of this matters if the end result - the cooling - is still the same as on 12V. Probably not, but is there a big difference? Let's check it out with a high performance Papst 80mm NGH (2.8watt) and a Delta ASB0812HH (3.0watt). I used an Alpha 8045 HS with AS3 on an Abit KR7A-raid. Both tests were done with CPUBurn, by far the most stressing program I know of. Tests started at idle temps, and temps were taken after 15 minutes of stressing (the temperature stabilized usually around 10 minutes). If the case temperature changed, tests were stopped and repeated when case temp was back to normal.

Papst NGH 80mm (2.8watts)

It seems like PCmods does a better job handling fans consuming less amps, just 0.5Volts loss. How does this reflect on a HS? Let's see:

Quite a difference! Pcmods handles the job very well while the Maelstrom falls behind with 2°C.