I've spent around a week living with both and I know wrist HR performance can be an issue for some on the Fenix 6.
Tracking overnight there is barely any difference
And this extends when both live wrist HR feeds are monitored in real time
For running, there is a quite big difference - the wrist HR performance of the F6 for running I have always found utterly useless, either reporting too low or cadence locking - on the run below the F6 stuck stubbornly around 15BPM too low for the first few minutes.
For outdoor cycling, then for me the 255 still requires an external belt - a hill sprint took me to 164 BPM with an external sensor on the F6, while the 255 never nailed it and got stuck around 149 BPM.
Both do indoor cycling fine - about the only activity the F6 can do reliably for steady state at least. I used a Polar Verify Sense in record to memory mode for this Zwift activity to cross check both wrist HR results - all 3 average 155 BPM with only 3 BPM difference for the max HR I hit.
For walking, the cadence lock issue seems to be solved on the 255 as well.
I would still always use an external HRM for the 255 but the HR seems improved during activity with arm movements like walking or running.
No idea if this sort of post is useful or interesting but just another morning coffee muse!