I think you're making a pretty big deductive leap here. I'll admit I'm a layman in these matters with limited software/electrical engineering background, but I'm struggling to understand why you're saying that it can't be due to a lack of processing power. I mean, there is a finite limit to the processing power of the CPU/GPU, so it's not unreasonable to assume that if enough cycles are tied up doing other things, it just cannot be done. After all, why would a nice smooth, post-processed visualization be a priority? Frankly, the driver doesn't NEED a visualization at all.
I'm curious what you're referring to when you say "this isn't news, this is what they've always done." Is this something that Tesla has actually stated? "Isn't news" and "what they've always done" indicate to me that there is authoritative data here.
Not to pick nits, but though they don't call the feature that describes the functionality of the current "Beta FSD" (i.e. Autosteer on City Streets), FSD, they do call the system "Full Self Driving Capability". There are only two options for Tesla cars purchased today: Autopilot, and Full Self Driving Capability.