With the project getting serious, it was time to move out the toys and start really figuring out what the heck I was going to be doing in detail...
Orientation Thought Process
I thought long and hard about how best to orient the machines in the space. My earlier stack setup had worked alright, though I found that I'd often not really pull out a machine as it was generally tangled in there pretty good. I also found that I'd forget about disconnected hard drives as I'd only ever seem them if I pulled the machines out, which usually entailed shutting them down first.
I then considered if I stacked them against the sides, I could have the machines on the right sitting upside down and keep all the motherboards against the outside walls of the space. This would let me maneuver (barely) between all the machines and change things without having to pull them out.
It occurred to me that in times I wasn't really working on the machines, I could fill the space between them by constructing a tall narrow set of drawers that I'd put on wheels. This would let me put all my cables and things between the machines and make great use of that otherwise wasted space from this configuration. And thus, I opted to go ahead and vertically stack the machines on the sides of the space.
Fan Selection Thought Process
I had a few people suggest that instead of computer fans, I instead use bathroom fans.. Enough so that I really checked it out. Below is basically how I ultimately came to compare things and make a decision. In all cases I only bothered consider air flow in one direction. So the below analysis is for a single span of fans (either intake or outtake).
I limited the bathroom fans I looked at to more than those listed above. Overall I found that in order to get the sound level near that of the Cooler Master, while keeping nearly the same flow rate, I would have had to have spent roughly 10x as much. Not to mention that the bathroom fans are physically much bigger and would have required substantial changes in to the ceiling of the space to accommodate.
Not having the power consumption information wasn't entirely inspiring. I bet it's printed on their boxes. But not even providing that information leads me to believe it wasn't a design consideration.
I ultimately opted for 3 Cooler Masters both for intake and outtake. The NZXTs are really nice, but after buying one and trying it, I found 37dBA to be rather loud. Alternatively the 19dBA of the Cooler Masters is nearly completely inaudible.
In my brain, the decision process was basically:
The highlighted rows above indicate the factors that stuck out in my mind. While the Cooler Masters overall cost a bit more to operate per year, their flow rate is basically the same yet their sound level is waaaay lower. Besides, I won't be surprised if they last a while longer than 3 and a half years. And thus, I also think the "Annual Purchase Cost" will be a less significant factor, leaving the fixed operating cost due to power, where the Cooler Masters are much better anyway.
And thus, in my limited analysis, computer fans win. With their quietness, low cost, small size, decent flow rate, they "win" in basically every way. I also preferred that given when I was putting them, powering them was a nothing project, as I just hooked them on to a rail on my power supply for my router.
** I'm not totally sure about the mechanics of calculating the total volume level. I'd think that if the fans all functioned precisely the same, one might be able to somehow run them out of phase (180degrees for the 2x case and 120degrees for the 3x case) and effectively have them cancel out the sound from each other. However, since I'm not sure they really produce a uniform tone as a sound, nor can I control them that precisely, I can easily imagine the group of them being louder than 1 of that type of fan. But I'm sure getting them to be linearly louder (like 19.00 dBA x 3) would be as unlikely as having them out of phase so as to cancel each other's noise out.