So, we know .2 and .3 have been deployed on cars both with and without radar, correct?Again, why do you think that would matter to if the car is using EXTERNAL RADAR or not?
You again seem to be desperately trying to move goalposts to cast doubt where there isn't any to be had based on the known/raw info from the internal look at what the system is doing.
But to reiterate another point Green made recently- only no-radar-hardware cars appear to be using the internal camera for driver monitoring as of the most current software (well, and possibly those still on the months-old FSDBeta software but that's even further afield of the topic).
Weird. 5 minutes ago you cared about interior camera or not, but now you're comparing 2 cars that both have cameras.
Anyway, any difference you perceive might be attributable to .2 vs .3, though both are still using radar.
It's not attributable to the model difference since there's nothing different in the software that cares about that in terms of perception or driving. (what do you imagine even COULD do so between a 3 and a Y?)
Is it possible that even on the radar cars, it's using vision primarily, but tracking radar values too as a verification to make sure the new vision values match the old radar values within a certain margin of error?
I'm not a software guy, and I have no proof of any of this. Just my thoughts from the info that I've seen floating around on the forums.