I also saw the effects of the liberalization of immigration regulations in the 1960s which saw floods of foreign students, many Indian, Chinese, African, and others, which had profound influences on US academic institutions, the communities around them, and, to be honest, on their home countries, one of which was a brain drain.
On that last point: one of the arguments for increased global wealth and/or population is that it would increase the number of geniuses and intellectuals (along with the rest of the population, one might suppose). That's possibly true, but I strongly suspect that the existing mechanisms for recognizing and promoting talent have already encroached on much of those possible gains. Not all, but a considerable fraction. Which means that the additional benefit would be smaller than a naive estimate would suggest.
Teaching opportunities are not good; there are far too many PhDs awarded compared to job openings for university positions.
It may be a benefit for private industry work, but remember that a PhD is primarily about showing you can perform rigorous scientific research, and compared to a Masters, matters less (or not at all) in jobs where research is not the focus.
My preferred reform would be for H1B to be fully unrestricted. If you're going to allow labor into the country, let it compete on fully equal footing with domestic labor. For the reasons Smith articulates and you've either failed to read or are intentionally choosing to ignore, employers already have the upper hand.
A better balance of power would be to repeal anti-unionization laws such as the Orwellian (100th anniversary of his death by the way) doublespeak "right to work" (really: "right to fire, and disrupt organization attempts") and ensure that labor has the same access to collusive collective action and mobility as capital and finance.