Yeah this whole mindset of 'conservative = accurate' when it comes to tornado ratings needs to stop, now. So many underrated tornadoes in the past decade (Chickasha, Goldsby, Vilonia and Chapman 2016 being the ones discussed most often on this site) that if not a single EF5 from this accurate then I'm giving up with the EF scale, or at least how it's currently administered.
I find it rather striking the way the lower bound is used as a default. The EF scale has three 'hard' windspeeds for each DoD, lower, upper and
expected. I have no issue with initially announcing a LB on first appearance (imagine the hoo-ha if a tornado was downgraded) but surely the rating should revert to near the expected value unless there is evidence in either direction (the building is better/worse constructed) as it's, you know,
expected - it should be the default. Yet the application we actually see is more like 'minimum until proven otherwise'. The upper bound for a house is 220 mph yet it seems to be very difficult to achieve even the 200 mph 'expected' without surrounding contextual evidence being astonishing. Either the application is poor or the standard of 'well constructed' for the
expected windspeed is so rare and difficult that it defeats that whole point of the Fujita scale, namely that all categories are based on a very common structure (the ordinary house). Or both.
The wider problem is they can't even decide what the EF scale wants to be. For earthquakes you have two scales - magnitude (the actual energy release, Mw) and intensity (the damage, using the Modified Mercalli Scale or similar). One reason given for disregarding doppler radar measurements (unlike that the measurements are well above the ground, an argument I'm sympathetic to) is that it's a damage scale. Now firstly, Fujita's own use of non-standard indicators shows he didn't think of the F scale purely as a damage scale. More importantly, certain low ratings have the justification that for a given reason (
aside from construction quality) the damage could have been caused at a lower windspeed (the mobile home impacting the house at Goldsby comes to mind). That's bringing windspeed back into it. So the EF scale is
not a pure damage scale either, and is in some sort of no-man's land, one with perhaps a bit too much room for personal interpretation.