I was there for the whale experiment. It ran from March 2017 to HF19, which was in June/2017. Well before most of the price run-up. Also, as I recall, the whole point of that experiment was to simulate the linear reward curve, which HF21/22 just removed because of community dissatisfaction, so it's not clear how informative that experiment really was. The price rally in winter '17 was also correlated with BTC at 20K, which I think was far more relevant. As this chart from Yahoo finance shows, the price rise seemed to have little to do with the whale experiment. The whole crypto market was up.
You're right about flags redistributing rewards, but that doesn't necessarily make up for the psychological effect that flags have on people. It's pure speculation to suggest that the positive and negative emotions are in balance. In general, people prioritize loss avoidance over pursuing benefits (coincidentally, I just read an article on this. See link #2 here ).
Anyway, that wasn't the point of that item. I agree that we can't eliminate downvotes without implementing a different reward mechanism to incentivize self-regulation of voting strength, and it certainly wouldn't be a good idea to experiment with that on the main Steem blockchain now. It'll have to wait 'til SMTs or steem-engine tokens (or some other chain) can experiment with other reward algorithms. I specifically acknowledged that I'm in the minority and downvotes are here to stay. The point of that item was to merely to suggest moderation as a norm when downvoting over reward disagreement.
Thanks for the reply! There was definitely a positive mood about the linear curve around the time of HF19, so you might be right that it encouraged people to power up in the early-middle part of 2017. Philosophically, I still like the idea that voting strength was exactly proportionate to stake, so it's a shame that people eventually found so many ways to game it.