The game-movie analogy is not very useful when it comes to product development. Movies are two to three hour non-interactive experiences which have had techniques and processes honed for over 100 years and really don’t have that much in common with games. Sure, elements of the presentation are similar, but there is a lot more differences than similarities.

If you want an analogy to use, big games are more like theme parks. They both provide huge interactive and explorable environments, and end up involving a lot of custom tech.

That said, game development is inefficient. There are a lot of reasons for this – the amount of quality middleware in the market is small (its a tough business to make money in), not-invented-here syndrome, or just poor planning. Finally, writing reusable software components is just plain hard – people have been trying to figure out easier ways to do this for decades in all kinds of programming.

I wouldn’t bet on a programmers and managers from the financial industry fixing product development, though. From what I’ve heard from friends in that industry, their development is just as inefficient and screwed up as games. The difference is the relative budgets for their software to revenues is pretty small.

A final irony: half of the recruiting emails I get these days are from prop trading firms — even post-crash. They are heavily recruiting game programmers — why? Speed. The difference between their internal apps updating 10 times a second vs 60 times a second is a lot of money apparently, and the programmers who know how to do that kind of optimization tend to be in the game industry.