1 person @ $800 a month or
800 people @ $1 a month
or 16 people @ $50 a month
or 40 people @ $20 a month

That would put an end to the gpu oligarchy   

Im up for it @ 16 people....

I might be in for as much as $100. Do we want to do this if we might have to stop (for lack of funds or whatever)? I don't want blocks to take an hour for the next 5 months...

Also is it even efficient to raise it a bunch and make it that much harder for ourselves at the same cost during the next period? Just thinking aloud here.

From a purely self interested standpoint what is the optimal rate at which to increase power to maximize coins earned? Max it and hold it there or ramp it up? Can we purchase any amount that we want or only exactly 128 GPUs?

I imagine a bitcoind server would handle automated payments to participating members?Can bitcoin handle subscription payments?
Maybe paypal or cc subscription might be a good way to manage the monthly cost .

Forgive me for being an ass, but if you guys do this, there will be someone else who will think "I want a cut, but not share it" and do this too, and probably a 3rd one will emerge, and my gut feeling is that there isn't enough to share amongst everyone. We will be generating 50x the notmal amount and, while difficulty updates are capped, it is sure to throw many more coins on the market than we expect. And you guys see the inherent value of bitcoins, but others will not and just sell all they generate on the spot (on the month bill cycle, probably) at whatever price they get. Great for buying, but still...

Having said all this, I'm in for $50 too, and can create an automated manager, but I lack the time right now.

Or use this for the stability of the network. Let me explain:

1) People donate bitcoins to buy this service.
2) Service starts crunching for a month.
3) Yielded bitcoins are sold in an exchange.
4) Money is used next month to use the service again (aka go back to 1).

Advantages of this system:
1) Even people with powerful computers can't control the network.
2) No hassle to keep track who gets what (if they want the bitcoins they buy
it with the money that saves them the next month, what they would have pledged).
3) If people keep putting money in this, it becomes stronger (enough money to buy
2x service, just do it).
4) Even not rich people can join.

Lets see, 30 days has 43200 minutes.
1 block (50 btc) every 4 minutes = 540000 btc.
540000 * 0.06ct = $32400

Man, that's a nice extra each month

There, fixed it for you

It will work for the first month, then you get 3 or four of these US$1k/month machinery going after the same thing, but difficulty needs to creep in and even the playfield. So after the dust settles you will have made a buck, a bunch of them actually, but once we're back to 6 blocks/hour we're in the 220k~250k coins generation per month, which nets you US$15k. Very nice, it you don't sell the coins... and there aren't a bunch of people doing it.

The difficulty isn't set by time, but by block count, so yeah, it would go sky high within the 2 first months making this pretty much impossible to scale... you would still get the whole of the generation capacity of the network, though, so as long as you can pay for the servers with ~200k coins, you are golden... but if you try to sell the coins to get US$, you'll be loosing value, so it's kind of chicken and egg, I guess.

$400 will get you 2x M2050 for two days. The free trials are hard to get as they're very popular.