I think the difference is that one of the data sets adjusts renewable energy upwards to take account of the fact that renewables deliver a greater proportion of directly useable energy - but I didn't follow the explaination any further than noting it.

I think it is probably true that both solar PV and wind have reached a cost inflection point where they are the cheapest avaible option and the previous 10 years are not a super-reliable guide to the next ten years. I'm just looking at the increases in deployment required and wondering where the factories to make the kit are being built today in order to expand production that much in the near future.