OTOH,
It is no secret that Tesla's estimates have slipped. Anyone who has researched Tesla would know this to be the case and would take the original estimates with a bit of a grain of salt. When I get forward-looking statements from companies, it almost always comes with a disclaimer stating clearly that the final product and general availability is subject to change. One should set their expectations accordingly when pre-buying future feature and function. Even Elon admits this and has been known to poke fun at himself from time to time as it relates to his delivery times.
I was happy to invest 2K into FSD (I didn't do it when it was 5K) because I felt it was a bargain. Tesla still has a way to go, both in terms of regulatory hurdles, but especially so technologically. They've collected an unprecedented amount of data on the roads from their fleet, and are in a unique position to leverage that data, but they've got to be extremely careful about how they go about this.
I have my doubts in terms of how far we can push the current sensor technology deployed on model 3 even with processing gains of HW v3. It is not my expectation that we will be able to take our hands off the wheel quite yet.
My feeling is that Tesla will need to do a lot of work (perhaps five years worth) to get to level 4 autonomy and maybe another five to get to level 5. If they come in sooner, I will be pleasantly surprised - if they don't, I won't be upset or angry.
Watching Tesla's evolution has been nothing short of incredible, and it is my hope that they continue down this road with positive momentum until they fully realize their goals.