Four reasons spring to mind for why this might have happened:
Equipment failure. I heard plenty of strong signals and received good reports so probably not.
Cross polarisation. I use an EFHW in inverted L so it probably radiates vertical and horizontal polarisation.
Multi-path signals cancelling out. This usually produces varying strength rather than a consistently weak signal.
Antenna orientation. My antenna wasn’t quite parallel to the wall and I think it was pointing roughly at White Hill. I would expect most of the radiation to be broadside so perhaps this explains the weak signals.
A weird one. Couldn’t really say any of your options were the likely cause. My equipment could easily have be the cause, I set up the 30m antenna specifically for the S2S, and swapped radios to one which covered 30m.
The surface mount components are already fitted, so unless you take them all off, soldering them on the summit is not an option.
Hans orders his boards with the SMD parts fitted at the factory I believe, so even Hans couldn’t provide a bare PCB.
The QFP processor would be a challenge I think, I’ve never soldered an SI5351 out in the wilds either.
I did once build a surface mount transmitter on Ingleborough G/NP-005 and I built a QCX on Great Whernside G/NP-008. I think building a QMX would be taking things too far.
Not least because you’d need a computer to install the firmware. Manageable, I’m sure, but one more item to lug up the hill. I, however, would need a microscope of sorts to see what I was doing…
