Sorry to hear that a brand new WH65 shows such behaviour ...
what is interesting for me is that the drop coincided with an extremely high gust peak. 139 km/h (measured by a WH65, meaning, it might have easily been 160+ km/h, as the WH65 is usually under reading wind speed) is also for a sensor array device kind of a stress situation I'd say. It's possible that, as this storm was accompanied by rain - so I read your post - some 100% humidity (=water) might have touched some of the ICs and might have caused this drop.
questions:
a) is it now continuing on this dropped level, proportionally ?
b) (a rather invasive option) is it possible to take the array in, remove batteries, put it in a dark room until the LED stops blinking (might take 2+ days) ?
Then put it on a heater or blow with a hair dryer into the radiation shield for 15-30 minutes - even open the radiation screen for the procedure.
And then put it back. As said before, in this extreme situation water might have entered the enclosure.
So maybe just a hard-reset is needed (that's why the LED has to go off and the super-capacitor fully unloaded).
It might go back to normal.
As for the array itself - I have mine (an I-shaped WH65, but technically the same as a Y-shaped like yours) for 2 1/2 years now - with gust highs of 106 km/h and rain exposure of 102 mm/h, -10°C in winter and +42°C in summer (temperatures inside the radiation shield, actual device temperature might have been higher), and it's just working normally without any "misbehaviour".
And I agree with Rover1822, that the probability two times failing the same sensor out of 7 sensors in two different devices - for the same reason - it very improbable (or there would be many more such observations if it were a systematic hardware failure).
Let's see. Maybe the drying and hard ("cold") reset of the array will bring it back to normal.