Interesting 2m/70cm WSPR path, Reflection & Diffraction?

I reported in the Forum that we have been seeing a very large and interesting difference between two very nearly identical paths on 2m WSPR. After about a week of 2m we have shifted to 70cm to compare. Perhaps tellingly, we also consistently see a rather large difference though this time only 10 dB but it's in the opposite direction. WW6D who with the same ERP and almost identical path had been 24 dB weaker is 10 dB stronger on 70cm.

I'm now entertaining a perhaps slightly crazy, maybe too simplistic, model for what my be going on. Suppose that both paths involve the large mountain directly behind and beyond KK6WUU, Mt. Tamalpais, as a reflector. Then suppose that an intervening mountain (near Mt. Burdell) is "poking into the path" and acting as a diffractor. People often throw out the phrase "knife edge diffraction" relative to VHF and microwave propagation even though rounded hilltops have been shown to diffract similarly to truly small, sharp edges. Still our experience might fit this idea.

I've attached a sketch of what might be happening. This is a topographical profile with exaggerated vertical dimension. In case you haven't found it there is a Nice On-line Tool useful for displaying radio paths. The two paths are quite close , they have no more than a couple of degrees of angular separation and the topographic profiles look almost identical but the vertical angles will be slightly different. But for an intervening ridge near Mt. Burdell, all stations would be LOS to Mt. Tamalpais.

I suppose what we should do soon is run a Quadcopter or Balloon Aerostat with antenna from ground level up a few hundred feet to see if we see alternating min/max SNR regions that would have to be present for this model to be confirmed. Maybe we'll do that one of these days..

Maybe someone else has another idea to explain it?

Glenn n6gn