Some misconceptions here.
Firstly, SPPT customization and modifying was never supported to begin with, nor officially "condoned" by AMD. So saying that they haven't ruled out further overclocking in the future means absolutely nothing since it was, technically, a "hack". When nvidia locked down their cards, it was for good. That's why everyone loves the 1000w bios', because it let's the cards stretch their legs.
2. These cards are absolutely not at their "maximum" out of the box. The 6900 xt came with 1.175v core voltage and 300w power (250 core I beleive) budget. Many people, including myself, have it much higher. I'm running 1.325v @ 500w max power. Core clock is solid at 2850 or so, as opposed to the 2500 or less you'd get maxed out on the stock 6900 xt. They most likely want to keep it locked down because last gen the tweaked 6900 xt was basically as fast as a 6950 xt, minus the memory bandwidth. This gen they can come out with higher voltage, higher power cards with much higher clocks providing performance you can't get from just changing a few parameters in a program or registry.
3. Cache, memory, fabric and core voltage are not the same and are controlled differently. If there was stacked cache on the Rx 7000 series, I could see that being a problem. As of now, it shouldn't be a problem.
Lastly, I'm sure some people will say "people pushing 1.3v or more" are part of the reason because it let's people kill their cards. It's easy to tell when a die is fried. It's been easy to tell for as long as I can remember. Secondly, someone might bring it that 500w is over-spec for the 2x8pin cables, etc. 188w per 8pin or 6+2pin is the minimum spec. Any high end, or decent really, PSU will use a gauge that can provide far in excess of that. Amps, not Watts.
I'm not saying this to be condescending, if it comes off as that I apologize. There are just factors at play here I think some people don't realize. I'm sure there are many factors I don't understand either.