lm1875/tda2050 inverting mode

Inverting and non inverting circuit are the same thing, just with the input node and the feedback ground node swapped.

You can thus always convert one into the other, with the obvious caveats re: input impedance and gain.
This isn't correct. The inverted mode differs mainly (apart from the input impedance) through the fact, that the input differential amp (LTP) turns into a cascode. Sonic performance is better in most cases - check out post #14 under
https://www.diyaudio.com/community/threads/non-complementary-class-ab-linearized-amp.202526/
and post #39 under
https://www.diyaudio.com/community/threads/the-rush-cascode-possible-wiki-page.137509/page-2
In case of a power amp input resistance is (as long as the NFB resistor isn't too high) between 1K and 5K and not between 33K and 200K as most usual for non inverted mode. Most papers claim, this is a disadvantage (and requires an additional input buffer) but in real life this is more of an advantage especially without additional input buffer (except for high output impedance preamps - mostly equipped with 1-2mA small signal tube).

Maybe one of the member can upload the mentioned Interview with either Mr. Bob Stuart or Allen Boothroyd (Meridian) printed German Hi-Fi magazine "HiFi-Exklusiv" (check out attachment from the mentioned URL).
After reading this everybody understand, why the inverted mode provide better sonic performance than the non inverting mode.

P.S.: Mr. Bob Stuart and Mr. Allen Boothroyd are also here in interview:
https://www.stereophile.com/interviews/906bob/index.html
more URL's
https://www.stereophile.com/content/40-years-meridian
https://restechtoday.com/audio-designer-allen-boothroyd-died/
https://www.av2day.com/2020/06/co-founder-of-meridian-audio-allen-boothroyd-dies-at-age-76/