Correct me if I'm wrong, but I do believe that even Tim Marshall has said EF5s are still possible with the current scale as is. I believe he used Jarrell and Joplin as examples of the damage he would agree to as EF5 worthy?
My only qualms with the Enhanced Fujita scale is as follows. As a scientific, damage scale, there should not be EF-U unless something truly did damage that is immeasurable. Qualified measurements of wind effect on an object or surface can be correctly, if not close-range estimated by mathematical formulas, even if it's not a DI on the current scale.
Likewise, Radar or Doppler Wind measurements should be considered within reason. It seems widely accepted that the windspeeds recorded on the DOW from May 3rd, 1999 are correct and measured the fastest winds on earth from the 1999 Moore F5. I understand ground friction and surface winds will not match those aloft where these have been measured. If something is on the line of High EF4/Low EF5 damage at ground level, however, and a DOW recorded say 240mph winds at 300m AGL, I would think that would be something to be weighted more towards EF5.
My final thought is this discussion on tree damage. Fujita ascribed the multi-vortex nature of tornadoes as an explanation on why major damage can be done to one area and little to no damage can be done 50-100 yards away. Unless it's something such as a question of debris load vs wind load, the spotty nature of damage shouldn't be a red flag on its own. .
Saying all of this, I do also understand damage estimates are being done within the realm of both established practices, established ratings, and what is shown on the current DI guide. This isn't a grand conspiracy by the ASCE or Tim Marshall or anyone else as much as a system with flaws that we have in current use. I'm awaiting the revised EF scale to see what ideas are accepted in the final edition and what is denied as a DI or measurement analysis when it comes to the end-all,be-all of a final rating.