It's obviously silly, but does it really sound like fun? My takeaway from this writeup is that these poor folks forgot to have fun.
I really loved the part where the heroic manager talks the Rails programmers out of embarking on an entertaining game of competitive hacking in order to keep the entire, gigantic 70-person team focused on... waiting around. Imagine the consequences if he had failed!
I think they were trying to combine two ideas, "startups are cool" and "bring a bunch of people together." But not all combinations of good ideas are themselves good ideas. This was a particularly bad combination, because one of the defining qualities of startups is that they're small.
If they had split the 70 into 30 teams of 2 to 3 people, and each team tried to build their own prototype over the weekend, some interesting results would have been produced. Put the prototypes online, and have the participants select the best one through a vote.
That would bring a bunch of people together without violating the "small startup" principle. Also, completion between the teams would help with motivation.
The way they did it I think 2-3 is too low. They have too many support types: lawyers, designers, business, public relations, etc. I think they'd have needed to have teams be 7-10 to fit that many types into a startup. Which is too many for a startup... so I think that retrying without the deadweight of support roles might help (not that lawyers/designers/business/public relations people arn't great, but its hard to perform their roles before there is a product in existance).
This is the first I'd heard of startupweekend. They seem to be off on some basic ideas, as other comments here concur. Throwing more developers at a problem doesn't usually solve it. A weekend is barely enough time to think about an idea thoroughly.
But it does sound like fun though. A barcamp should take the idea to the illogical extreme and hold a startup-hour session. :)
