Does the comment on line 153 of the .pio file need updating?

 		.program pio_i2s_in ; Note this is the same as _out, just "in" and not "out"

Not to my mind. My comment was that the change was "only for input mode" because the output mode already has the correct timing.

I'd be happy to make that fix, but.... can someone confirm that swapClocks is working for inputs? It works fine for I2S outputs, but appears to have no effect for inputs. The bits look properly swapped in the PIO input code, so either it's in the C code or I'm doing something silly.

Not sure I understand the question, sorry. The "swap clocks" versions here refer to just changing the order of the BCLK/WCLK in the PIO sidesets. So 01 in normal mode becomes 10 in swapped. It's not actually changing the protocol, just swapping the BCLK and LRCLK in the hardware pinout. So a line-for-line copy of the fixed version, but reverse the sideset pins...

You're right; it's a trivial change. The problem came when I tried to test it and discovered that, for whatever reason, the clock-swapped PIO code wasn't being run. Just to make sure, I then ran swapClocks() when setting up the output I2S, which did swap the clocks. So either I'm doing something wrong or there's a bug in the library C code. If this isn't an open issue, I can dive into the library code and see why it's not calling the right PIO routine.
Meanwhile, I can also just make the literal 2-bit change in the PIO code, with the understanding that I can't test it in vivo.

Yes, please do the change. We can debug the swapclocks bit elsewhere. Thx!