Re: ATSC 3.0 fixes a lot of problems with ATSC 1.0
In fact, if you look at the research both in the lab & in the field, noise & multipath rejection of 8VSB is in most cases on a par with (within a couple of dB) the 64QAM used in DVB-T, and slightly better than that found in 256QAM in DVB-T2. And, despite early concerns, they managed to retrofit SFN operation to it fairly successfully; although not used much in the US due to the market structure there, the few SFNs they do have are comparable to the largest DVB-T SFNs elsewhere.
If they really wanted to improve those factors that much, they would've adopted something similar to the ISDB-T used in Japan, etc. Better noise & mutipath rejection, more flexible SFN operation, hierarchical structure allowing for low power ESB/EW receivers, etc, etc.
In reality, it seems ATSC 1.0's near-complete lack of extendability doomed it.