"works well" is very subjective, you need to be a bit more specific.
So far for me it works about the same as any other GPS watch that doesn't do speed source from footpod, i.e. it has a definite lag on both auto-pause and auto-resume, usually at least as much as needed to cross the street and often a bit more. Very much the same as I experienced with Fenix3 and with 920xt before the latter go the speed source from footpod which made autopause much less laggy. I'd returned the Fenix3 before it go the same feature.
Sorry for the late reply all and happy holidays :)
As for my findings, when I'm riding my bike and I come to a stop at a stoplight (or even a stop sign), I get an immediate vibration that the activity has been paused.
When I'm going for a run, if I come to a stop light and even wait a minute or so, the activity is not paused and still recording (and my pace dropping as a result). Perhaps I should get a Garmin footpod so that my speed is better recorded instead of being derived from the GPS?
I have not tried custom pause option yet since well I have no reason to pause a run after a set duration of time.
I believe the autopause setting is activity specific. Are you sure you've enabled it for BOTH the Cycle and the Run activities? Your description sounds as if it's not enabled for the Run activity type.
The custom pause option defines the speed/pace at which autopause kicks in in case you want a different threshold speed/pace.
Afraid a foot pod outdoors won't help - as there is currently no option to use foot pod for pace but still have a GPS track. Don't know if this will be added in the future. It was for the 920 eventually as you mentioned.
Also why do you think your cadence needs to be better tracked? The watch does cadence just fine in my experience.
If Auto Pause is laggy and, this is really important to you, another option is just to stop the timer if you have to stop and restart it when get going again.
As I said above I don't worry about this any more as GC or Strava computes moving time for you.
I've read an article on fellrnr that the internal accelerometer wasn't that accurate but nevertheless my main reason is for indoor running.
But yeah it's not a super big deal, I pause my workout (which is time based) during stoplights anyways. I just wanted to know if it was an issue or not for others.
I guess there could be an issue with cadence from the watch however if you stop moving your arms for significant periods. That's certainly true if you are using "WDR" pace (from the watch) indoors but never seen much difference with cadence
