There used to be a much stronger push for education in the US.
Perhaps if companies could not "just hire from overseas" or "just outsource" there would be a longer term growth strategy that would focus more on the education of the US population (not just training for this or that job).
It did seem in the past that there was much more of an all-hands-on-deck attitude towards education throughout US corporate activities, more broadly focused on the general fields the various companies valued the most. I suspect this fall off is very real, but don't actually know if that is just my impression or if there is a concrete effect from modern economic structures.
It's an important enough question it should definitely be studied and taken into account in policy.
However I can't agree with your conclusion that "Immigration helps the countries [sic] top-line metrics, but it rarely helps the citizens inside the country". That requires meta studies that I have never seen to prove it is so. I could cautiously accept that "some types of immigration rarely help corresponding sections of the local population" much more than such a blanket judgement. Overall, it is just not true that economics is zero sum. It doesn't have to be. An entire people can in fact flourish.