This article at Broadcast Engineering is from a while ago (October, 2000), but it is another interesting point in the argument of whether 8VSB or DVB-T (COFDM) is better for amateurs.
In short, the peak-to-average ratio for COFDM is at least 2 dB greater, and the receiver needs about 4 dB more signal to receive. In other words, you need to pump out 6 dB more signal for the same coverage - that's 4 times more power. The article also mentions that COFDM is more sensitive to QRM than 8VSB. In the past, 8VSB used to have a harder time with multipath, but modern receiver chipsets have come a long way in ghost cancellation, and have largely caught up with their COFDM counterparts.
The reason we have a different standard of broadcasting than the Europeans do is due to the very different nature of broadcasting in Europe. In general, in the different countries in Europe, they have a relatively small number of broadcasters that have nationwide coverage. It would be as if NBC, CBS, ABC, Fox and PBS all had one channel that was replicated across the entire country with no local programming of any kind. The broadcasters in a particular country serve the whole country with multiple transmitters, obviously. With the advent of DVB-T, design decisions were made (by picking COFDM) that allow for multiple transmitters to operate on the same channel without interfering with each other - a so-called single frequency network (it's not really a single frequency as much as a single channel, but that's just a minor detail). Here in the US, Each individual broadcaster has a single transmitter (yes, some of them have translators, but those are special cases), and those single transmitters cover a much, much wider area than the individual transmitters in Europe. For US broadcasters, the higher peak-to-average penalty and the reduced range of power of COFDM would have been a bitter pill to swallow - particularly given the fact that broadcasters here actually pay for the spectrum rather than getting handouts from the government. Given the different reality of how broadcasting works here, it's no surprise that the FCC chose 8VSB.
I mean, to my mind, it's still moot: If the North American market had thousands and thousands of COFDM receivers, then that's what I'd be planning to use. The decision to go with 8VSB in the broadcasting universe was, for us amateurs at least, serendipitous.