>Maybe the goal was to get more people who could quickly assemble a MVP from existing APIs that appealed to the emerging cohort of buyers and trendsetters.
There is a lot to be said for building the right thing; but the actual ability to assemble such products are very common and even the very premium skills are definitely for sale at market rates much lower than $15MM per year.
The skills are common here, not in general, and $15MM (total, not per year) is a cost of purchasing the assembled infrastructure, not the idea and potential to build. I don't think this is a wise expenditure, but you're attacking it from a poor angle.
There is a lot to be said for building the right thing; but the actual ability to assemble such products are very common and even the very premium skills are definitely for sale at market rates much lower than $15MM per year.