Sx remove delays

[jlpoltrack]

Remove delays for SX126x and SX128x for functions used in main loop. Note - LR11xx never had these delays implemented.

[olliw42]

many thx
looks sure reasonable, but some historical cruft of hesitation
when I started doing this for the sx128x, the semtech driver had some code to ensure certain deadtimes between certain commands (also elrs did that). You can find the function SetDelay() which is there to do that. Very early on I had such code, but over time this got less and less as it seemed to not hurt. Eventually, I didn't use SetDelay at all anymore, but hardcoded some delays in "strategic" places. The traces of that is what you see here.
Having said that, I think we need to carefully test this.
Unfortunately, I do not have a good idea of how we could do that besides just running plenty of hardware combinations for long times.

@brad112358 @tmcadam any ideas? any concerns?

[jlpoltrack]

One more note to add here - I'm pretty sure mLRS won't be communicating with the sx chips for some time (ms) after these commands.

[olliw42]

I'm not much concerned about the normal operation but the edge cases e.g. in low connection conditions, ehere we may get the more unexpected conditions. I feel we jsut don't know eneough of teh behavior then.

Proposal: we keep this open for now and try to move as quickly as possible to v1.4, and once v1.4 is done we merge that immediately. Gives us time to watch out for issues until th enext pre-release. I mean, we also will have to do the swicth to gcc13 at some point, right after v1.4 looks like a good point for all the "risk" changes.