Square waves were requested, so here they are:
It was also pointed out that having the speaker negative binding post connected to audio ground would bypass a resistor in the feedback network, and this is quite correct. As suggested by the person who brought this to my attention, I left the load and my scope (running on battery power) across the binding posts and grounded the feed to my FFT to the audio ground instead of speaker negative. This change did not appreciably change any of the readings I checked.