I now own a spectrum analyzer! I bought (and then it was paid for as a birthday present) a Sencore SLM1456CM. It's a very nice tool, and has been extremely useful already.
One of its features is an ability to find the SNR of digital signals. (The term is MER -- Modulation Error Ratio.) The analyzer tops out at 36 dB MER even though the number can go higher in reality. 16, 19, 27, and 41 all have MER values >36 dB. In my observation of 29, I have not seen a number higher than 32.9 dB, which is still much higher than the FCC spec of 27 dB. Plus, it could just be multipath or overload on my end.
WVPT-11 has been a pain to catch and I've been in communication with them. The highest number I've seen is 22.1 dB MER, but I assume that's just due to poor reception. I figure a proper roof antenna would provide a higher MER, but as much time as that signal spends lost in the noise, I can't imagine most people having a lot of luck with it. (Amusingly, I had really good luck with WVPT today while in a grad student office using a paper clip as an antenna. That's where I got my 22.1 dB number.)
Just thought I'd let everyone know.