> I think there's a strong argument that RCV/Approval would shift the incentives and game-theoretic landscape away from polarization and turnout, and towards big-tent coalition building, to appeal to the greatest quantity of left, right, center, and independent.

I think most people wildly overestimate the popular appeal for compromise. Only about 20% of the population would support an opposite-side moderate over a near-side extremist. A true centrist party is doomed to lose in any system that knocks out unviable candidates, since it's the first choice of almost nobody.