That works for a very short war, like those in Iraq/Kuwait in 1990-1991 and in Serbia in 1999 happened to be, meaning it was cost effective to have invested so much money in the far more superior US planes because the wars in themselves were short (so the US got a good ROI out of using those advanced planes).
But what to do in case of a war that spans over an entire year or more? (like the current war in Ukraine). The marginal utility of those first debilitating air strikes carried out by advanced planes goes down quite rapidly, nobody cares after 12 months of intense on-the-ground war if some airfields behind the front have been blasted just as the war started, most probably those airfields have already been repaired.
And there's also the opportunity costs. The money spent on developing a very advanced plane like the F-22 could have been put to better use by investing in the US's artillery shells-making industry (of which the West has not enough of right now), artillery shells which provide a much better ROI during a 12-month land-war compared to an advanced plane.
It's all a matter of what future wars the US plans to get into and how it thinks those wars will actually develop.
You then use your vast fleet of less capable planes to ramp up sorties, provide air support during combined arms operations, etc. You can still use your magic airplane for protection of those assets during sorties, and for strike missions on high-value, high-threat targets.
That is why it is not a problem to have 'few' expensive magic planes. The US only had a hand full of U2's and SR71's, because you're not doing spying runs over Russia 10x a day.