I bought it, but I bought it as a donation towards funding R&D in the space of autonomous driving. I’ve found FSD useless outside of highway driving, which adaptive cruise control does ok at. That has however changed in the last year and the FSD has improved dramatically over months to the point I can reasonably navigate door to door in a hostile and negligently adversarial driving environment like seattle without fear of imminent death. If I were evaluating it on “what could I have materially attained for the price of FSD” I don’t think I got a good deal. But again, I didn’t spend the money with that in mind. I bought it with the idea that $3b in consumer crowd funding of FSD will materially advance the end of human intermediated driving - which I think is akin to curing cancer.

I think it heavily depends on when they bought it given the increase in pricing.

I bought it at $7k, knowing full well that I’d never likely see true FSD, but the features it added were worth it to me like navigation on autopilot, and lane changes on the highway. We figured that just those two were enough to spring for it, given we commuted a lot and did road trips a lot. So it was nice to have it take over every once in a while.

At $7k, it was pricey but not very out of line with other car companies upgrade packages.

I think the ROI really depends therefore on what price point it was got at (someone who got it at $3k probably likes it a lot more than someone who spent 10k), and what they were expecting out of it.

I think true FSD is a smoke and mirrors trick that Tesla will not achieve with their current sensor loadout.