Fascinating! Running Stockfish at a greater depth after 10...b5 seems to struggle to show any e5 suggestions amongst five lines at Depth 26. I wonder how far one would need to let it run to show a different result (I don't want to overheat my laptop)?
> So, can we trust the analysis???
For the most part, probably (provided you understand it and can make use of it) - unless you're a Super GM (who might opt for alternative strategies when playing others), or Leela Chess Zero, which is currently taking on Stockfish in a TCEC "Superfinal": http://www.chessdom.com/stockfish-wins-the-premier-division-lc0-qualifies-for-the-superfinal/
As I type this, they've played 43 games and are tied at 21.5 points each with the 44th game appearing like it's heading towards another draw. 56 games afer that, and then shorter time control tiebreaks if it's still tied...