Let me get this straight.
Your claim is that if a human beats you with a dumb strategy, that beats your strategy ~10% of the time, this should somehow be different from a bot playing a dumb strategy, and beating your smart strategy ~10% of the time.
Let's suppose you have a human that plays exactly like the bots. This should affect your rating the same way, since to the rating system there is no difference between the bot and the human's play.
You also say that you could be playing against somehow who is rising, and has not reached their true ranking. The inverse is also true. You could be playing against someone who is considerably better, but is on a losing streak and is underestimated by the system. The only difference between playing a bot and playing a human of the same ranking is that the bot is consistently bad, whereas the human could be better or worse than their ranking suggests.
So, if you strongly believe that playing games against humans makes your ranking rise slower than against bots, then you either believe that there are a considerable amount of rising players, or that the Goko ranking system is broken and somehow counts bots differently. The latter seems very unlikely to me, so it comes down to how many people are underestimated vs the number who are overestimated, and my guess is that after around 4000, there's not too many people like that.
(Personally, I actually had some trouble getting to around 4000 just by playing bots, because I was playing Base Only. It turns out my Base only game is garbage.)