I don’t think bot evaluation at the beginning really means much to human players. 0.3 point on 13x13 is just too small for us to care about when we constantly blunder 5 points away. As a comparison, I myself have played a few 9x9 games where every move is optimal (confirmed by katago).
Also, 13x13 is so big that nobody even knows if bots can find optimal play on it, which means OGS bot evaluation isn’t fully convincing. On the other hand, despite being very far from solving 9x9 completely, we are pretty certain a bunch of 9x9 lines found by bots are optimal for both.
Ok, we are very off-topic. If anyone is unhappy with it I can move to another place.