The flow sensor issue is really slowing me down.
The thread is BSP as you said, which is extraordinarily hard to find in the US. I've come to the conclusion that the fitting, and the way it seals, with the two smoothly machined faces separated by a gasket, is really a good system, but it doesn't seem to be employed in the US.
Mikesolar made some very relevant posts
HERE, and
HERE, that have given me sufficient information to find out just how difficult it is to get these parts, here in the land of the free, home of the brave, blah, blah, blah.
On your joint, I think that you could have eliminated part "C", and just connected "B" directly to your sensor with a proper gasket to seal between the parallel flat faces of your sensor and what ever it was that allowed pipe "A" to be tightened down with nut "B".
But anyway if the failure rate you are experiencing is typical, it really makes me wonder...
I did find an ultrasonic flowsensor
HERE that does look very interesting, no moving parts, it also looks like minimum flow restriction, but the cost might be brutal.
More later...
-AC