Let's assume that we plan to develop a new product, eg a new high power converter.
Sure we are going to follow a "gate model", a list of "must be done" milestones.
Each milestone has a budget and a time line.
Here we stress the importance of the "concept phase". Elaborate and go to details during the "concept phase", invest an additional 0.5 MY (for a 10 MUSD project) and cover risky tasks in advance. We know, this will cost money but will avoid "u-turns" in later project stages, and this will save huge amounts of money.
Making development "à la carte" is the right way to do the job. Some will claim that this is a bureaucratic method but trust me is the right method that will save money and increase the chances to get a good outcome.
And the most important, have fun during the process and balance carefully between "gold plating" and "fast and dirty".