Game Analysis Question

I was looking back at a game analysis and at one point, where I had made a mistake (as per the analysis) if I followed the suggestions instead it would have ended in the loss of my queen very early on in the game and suggested that it would have been the best move. Can anyone help me understand why? To me this suggestion makes absolutely no sense why it would suggest trading my queen for a rook as the best move to make, as the knight on B3 would take my queen immediately. Is the analysis actually trustworthy?

Notice that your rook on b1 is in danger from black's queen on f5.

This means that if instead of Qxa5, you instead move your queen somewhere safe like to b4 for example you will lose a whole rook.

public shaming is not allowed, please report instead even if they're clearly hacking

Be7 to attack rook on f8 seems to be the best ply, since qxRb1 would end up as a trade of queen for a rook, a loss of 4 points. So why the particular analysis deems Qa5 as the best move to me is befuddling. But I may be overlooking something.

I'm guessing that if an excellent ply is only 99% of the best move then the engine will calculate another as optimal. I think Bobby Fischer was right in lamenting that creativity in chess had sunk to third place, after preparation and memorization of patterns. Maybe we are being herded into playing like robots. Is the satisfaction of winning replacing the element of fun?

Also, after Qxa5 Nxa5 you can take the bishop on c1, so you're getting a rook and bishop for a queen which isn't that bad (only about 1 pawn difference)