I was something of a skeptic about the efficacy of power levels on high-VHF. Here in the Dallas-Ft. Worth DMA, WFAA increased just a tad from 45 to 55 kW in April 2010 (less than 1 dB of difference). Theoretically that should have been negligible but the real world worked out quite differently.
As one metric, a television equipped with non-amped rabbit ears here some 40 miles out, went from an occasional lock on WFAA (RF-8) to a solid signal 24/7. No change whatsoever was made here at the receiving end. That slight boost in power by WFAA was evidently enough to get past the digital cliff and provide more power density.
As far as I know, everyone in the whole DMA that had similar problems also benefited with that smallish 10 kW increase.