The current extreme polarization in the US is new and fairly unusual, in historical terms. It could be an inevitable phase that other counties will reach too, or it could just be a random artifact of the US's history and political system. Very few countries have a voting system that locks in the main two parties as strongly as the US, for example.
The idea that all politics is purely about tribal identity, and that "rationality" is a myth, would seem to suggest that we've never made any real social advances. But I'd count things like the outlawing of slavery as advances.
It reminds me of the extreme view of social structures in science (which Thomas Kuhn subscribed to, if I remember right), that all scientific "advances" and "revolutions" are just changes in fashion as one generation of scientists succeeds another. But each advance does in fact get us closer to the truth, even if we never quite reach it.
Without getting into the whys and wherefores, my main concern is that as we eliminate some forms of tribal clustering (ethnic, gender, class, etc.) we are simply replacing them with other forms, namely political affiliation or wedge issues. Particularly, I'm concerned with tribal clustering on wedge issues that don't affect the quality of work a person might do.
And in addition, the more and more fiercely we tribally associate, the less and less we want to do with the other tribes. This has gone to the extent that certain fields of employment, or even fields of science are largely owned on tribal lines.
Conservatives largely outnumber liberals in fields like agriculture, while liberals dominate social sciences. This sounds fine, but peer review demands cultural diversity as well, or our own cognitive bias imparts blind-spots on the work. The only fix for those blind spots is cultural diversity, and the only way people can be happy with that cultural diversity is to re-tie their Gordian knots and learn to accept other people as other people, and not the devil monsters we tend to equate them as today.