Its impossible to avoid the late Fred Brooks, if they spent $1B the due date would have been X, and spending the next $35B made the due date later.
For a given technology you can only sensibly produce a product of complexity X, beyond which it scales at some awful exponential rate, so throwing more labor at the problem rarely works because the organizational coordination costs scale up faster than the productivity gains.
Some comment I read recently (but I can’t find) put it in terms like “Meta spent 36 billion to make secondlife for the wii”