Another new one I have seen for the first time. See link
Hope the link works

Coospo s10

https://a.aliexpress.com/_mPCb6wd

I think this one is interesting. Apart from the temperature range the specs are extremely similar to the Thinkrider PP5.
Personally, I have been extremely satisfied with the (data) quality of my Coospo H9Z. It has been way better than a Polar H10 I once had and any other HRM I have had.
In Coospo’s comparison against the Favero pedals it looks like they did some sort of sprint/acceleration with a decent result. In contrast to this is the data example from the Thinkrider PP5 page. That one is done pretty much as a steady state effort and that shouldn’t be that much of a challenge to get right.
Okay, Coospo does not show any values for axis, and there most certainly is no guarantee that the data is true. I think the recent GPLama reviews have shown us that most manufacturers own data and testing cannot be trusted.

The front chainring position hypothesis is in line with my experiences. I am inclined to say that chainline also plays a role. My deviations become larger when I am in the small chainring in front and a large sprocket in the back. This could perhaps indicate that the power meter interface is not stiff enough.
Attached is part of a triple recording on Zwift with my Kickr V5, Favero Assioma Duo and Xcadey Xpower S. In the large chainring the deviation is close to acceptable (This until about 5540 seconds). Then as I change to the small chainring and a low gear in the back for the Radio Tower KOM the Xcadey goes from reading about 1% lower to 4% lower. A calibration at the top in the big chainring and the Xcadey is back to constantly being 1-2% lower. This pattern has repeated itself every time I have done the Radio Tower KOM and that at a low power output of about 300W.