That probably wouldn't work out economically viable for an AAA game.
The main constraint on players these days isn't cost, it's time. And if you have a ticking clock in the game with a price meter, they're going to feel the time constraint and be likelier to quit playing sooner rather than later. Notice that most games have a more time = more rewards model. This pricing scheme would run opposite to that.
Also, if you played for 40 hours to complete the game's main story and at that point, you'd spent $60, then what percentage of players would actually make it that far? I imagine the curve for play-hours would be very steep, with most playing only a very short time and only a handful of the starting group of players making it that far, and thus those people having to foot the bill for almost the entire development costs. Plus, it penalizes slow-and-steady type players, or those who look in every nook and cranny, so you've just alienated a whole class of player types.
Breaking the game up into purchasable chapters would be a better option. Pretty much anything that makes games less-monolithic is the way they're moving forward with that issue. For example Call of Duty's game-every-year model means that development times are shorter, each game has less scope to develop, and it's not all eggs in one basket. So you have the engine team working on the next engine version while the designer teams crank out stories and games using ever-tweaked versions of the last engine. It also means they're not as buggy: they're not shipping their latest game on the bleeding-edge engine, they're shipping games on a rock-solid tested engine, while doing research on engine++, and they can back-port improvements into the working engine after they're fully tested, if desired.