The one thing that always gets me in the variance of opinion, is general the BFTD person goes back in time a picks out a historical period that suits/backsup their assertation that BFTD is better. Then assert what they would have done to get all the different dips. But in a future price pattern is hard to guess, and really know are you getting it at the correct dip price. When we hit the ATL this cycle, I remember alot of people were waiting for <12k when we are about 15-16k mark, and probably missed the boat. The dca'r would have got some really close to it for a bit. When the market started rising I would be fairly certain the dip buyers probably got in at a higher price some where in the >20k. I think this scenario plays out quite a bit more than people are willing to admit during all cycles, and price volatility in bitcoin.
For the experiment the future market is a complete unknown, for both the dca'r and the dip buyer. It creates a level playing field and simulates reality, I think the experiment would help show some of WCS(woulda,coulda,shoulda) side of dip buying that you dont really have privy too.
That sums it up pretty well! The market is unknown and while there was a time when big players like Binance probably could manipulate the market if they wanted to (or did?), it's now a point where even the big players most likely have a harder time to really tell where it goes as the number of big players is growing. Unless they collude, it' more a question of tendency than price points. And as you said, nobody knew whether it would go down or up at at the 15-16k mark and yet you find all these gurus with their moving averages and what not to tell the world where BTC goes next. I have long, very long given up following anyone of these gurus. I can't even tell when I watched someone drawing lines the last time.
If anything, the only person or entity to know whether BTC goes tremendously down is the one sitting on at least 50k BTC and decides to dump. But that person won't let the world know, so no guru can know. You can twist any data points for so long such that any drawn line and any other indicator for that matter, makes sense somehow, at least somehow, and then propagate it as if it was genius analysis.