I think it's irrelevant if the cost of solar panels goes to zero, the reality is that the farm to fork cost/kwh is still quite high because dispatch and transmission are never cheap.

Furthermore, for every 1 pound of polysilicon produced...you get 4 pounds of silicon tetrachloride output.

I gambled on $UAN and $AMR guesstimating that the spread of renewables would lead to more nat gas and coal/steel consumption per a kilowatt hour produced globally. I got lucky and it worked out. I'm not bullish on solar costs going below the embodied energy cost of desal per 1000 gallons. (10-14kwh/1000 gallons)