Is this my heart going mad, the watch or the strap :-)

Very weird.
Two things came to my mind:
1. Battery of the hrm-dual is dying.
2. You had a connection drop and this HR peak was from the OHR.

There are some tools you can use to analyze the .fit file to check if you find something weird on the data. I don’t remember the name but someone here probably will help with that.

My take: I expect the sensor module to have changed in recent firmware updates and they try to patch things with different algorithms now. Anyways as a softare developer myself I think some quality assurance issues should be addressed. With the right testing approach, things needn't have escalated like this.

Everybody was focussed on Optical Heart Rate, but from this and other posts it is obviously not just the OHR (as I experienced change in behaviour with HRM Pro).

Another reason might be that the watch due to e.g. increased functionality has too many tasks running concurrently (could also explain battery drain) which could lead to either missing out measurements or lagging measurements (depending on the architecture used for asynchronous measurement intake.

Since not just forerunners have heart rate issues I suspect the sensor module. I don't suspect it's a hardware issue but it's a software flaw the way I see it and it should be traceable to release 16 or 17 I think. Put 2 developers on the changes in the version where it was introduced (and take all modules into account) and they'll find the answer.

Congratulations! You have successfully installed SW V 18.23! It is brand new and if it does not work properly (OHR+external HRM - ha-ha) you get a Rope Jump to be absolutely happy. 

PS Yes, of course, this is sarcasm. What else can you say when a manufacturer abandons its customers with defective devices and leaves for the Easter holidays?

I add for those who claim that only those who complain write here, that on the Italian forerunner forum on Facebook people also complain about heart rate