I'm not a believer in the Efficient Market Hypothesis, but there is at least some truth to it. The market has some ability to price in known future events such as reduced new supply. At this point we've already had two Bitcoin halvings. It's hardly a secret to only die hard bitcoin tech geeks.

I like your main point, supply is just one side, demand is the great unknown for Bitcoin. Can't just assume constant or even increasing.

There's also the fact that the supply impact is going to be much lower for the third halving. Since it only affects new Bitcoin, there's far more of it already out there and newly created Bitcoin is a smaller and smaller part of overall supply. Thus halvings should have successively less impact each time.