Anyone STILL having heart rate accuracy issues with their Fenix 6?

I contacted Garmin Support (again) with the problem, and they gave me the generic response (again) on "why the wrist based HR monitor" may be inaccurate. Basically accusing all of us of not know how to wear the watch or what to look for...

This is my first garmin product, and it will definitely be my last. Unfortunately, I don't have hundreds of dollars to drop another watch...

No tricks. Indoor cycling activity. Tightened up a notch over normal use and sat above the wrist bone. The ‘trick’ with cycling is to ride on the top of the bars or hoods and avoid gripping the handlebars tightly otherwise that will constrict the blood flow and make the WHR inaccurate. I raced a duathlon last year riding on the drops. In that position the wrists were flexed and I was gripping the bars. The wrist heart rate trace was pretty crap...but expected. 

Nah. I can do a perfectly unflexed, still ride sitting up, on the hoods, in the drops, arms crossed, extended overhead, etc. Doesn't change a thing about the crap capture. Then I go outside and ride a super technical dirt descent on the drops of my gravel bike with massive vibrations and it's just fine.

It would be one thing to have slight inaccuracy, but it's just wildly wrong, often spiking downward when I start an effort, then increasing again during a rest interval, as if there's an inverted conditional somewhere in the code, or it's dropping 'frames' when overwhelmed.

If it was fixable it would have been fixed by now. About time Garmin hold their hands up to this and offer people refunds or replacements when the issue is fixed.

WHR is new to me. It's strange I didn't had any issues with AW5 in any kind of sport however I wore it. Anyway, I can understand Garmin feels more sensible and also I can attest that using on more than 30min ride indoor on the trainer, I feel more numbness in the grown and hands (especially left one where I also use the watch). Most probably because of very fixed position and demanding workout to keep power and sometimes even cadence, compared when just riding outside where you move more to get the environment around, you have bumps, and so on.

I can understand also that for best results I can use chest strap when I really want the very exact values.

What confuses me, is that, as many observed and said, the measurements are consistent depending on the activity. I think they just try to fine tune the values, based on activity type, movement sensors and other things they can rely on. The thing is that maybe this is such an unexplored and variable territory that it can make more harm than just having some raw data from optical sensor. You can't just think of any patterns an athlete will use the watch. I've seen people here complaining this started after some firmware 3.0 or something like this. Maybe then, they started to make this, supposed to be state of the art technology, but it's still worse than anything else.

Again, I wouldn't complain if for best results, I need to wear the strap, while I can be assured that day by day measurements are fine and all the nice things I see in Garmin Connect are true, because otherwise, then they are just useless.

Basically yes, the AW4/5 have impeccable WHR.  Best in the industry.   But worst in the high end fitness watch for battery life  :).   I have to have 2 AW's.  I use one for dress and one for sport so that one is always charged.  

The Fenix 5 is looking good right now.  It uses the old sensor layout that works great!

They say the sensor is state of the art which is evidently not the case. Cave art perhaps, although even some of that is quite good...

The sensor is capable of very good performance. Unfortunately it seems to be hampered by software engineering decisions that produce stupid results. This may be in the pursuit of longer battery life or maybe other factors too.

For example, using Trail Run the wrist HR has kept pace with my chest strap up to 175 BPM. Putting in similar effort using Hiking the watch sometimes stalls at 120 BPM (this seems a significant threshold the watch strugglers to cross at times) or some other value well below the truth.

Outside of activities the watch is simply not geared up to deal with high heart rate numbers. It simply cannot register them. It's not that the sensor can't deal with them, but the software is artificially crippling the watch with an insufficient sampling rate that can't keep up. I've reported this to Garmin, as have many others, and it seems they don't care. They admit the fault, or don't deny it, at least, but it remains unfixed. So a design choice, even if one that no end user ever asked for.

A run with two F6s last evening. It's a lot better than some folk get but the difference is why I always wear a chest strap for 'active' activities as Garmin recommend.

The F6...