I disagree, somewhat. A million dollars a year can be a lot of money. But it's very difficult to constantly maintain spending that money in the most efficient w

I disagree, somewhat. A million dollars a year can be a lot of money. But it's very difficult to constantly maintain spending that money in the most efficient way possible. And all it takes is making one big mistake to lock in the development trajectory to a track that is very inefficient and suddenly slows down with the given budget. At that point it becomes very difficult for most organizations to fix things, because it means abandoning work that the team has already planned for, partially completed, made commitments on, and so forth. The only way to get back to the more efficient model is to make a drastic hard-break and abandon a lot of current work. Though it might be very non-trivial to recover the state of development back to a previous point. Such an effort is also likely to be very demoralizing as well. The trick is to have someone in a position of power making sure that such "trajectory excursions" either never happen or are easily recoverable from.