Energy transfer from a compressible gas is a complex equation, but you can stick a V cubed component in there and not be ridiculed.
Maria had 160 sustained, at Dominica landfall, compared to Irma’s 185 on Barbuda.
V cubed in that comparison, would equate to 1.56 times as much energy.
Thing is...wind related structural failures are all about openings, and getting the roof tied down using steel in shear, not steel or wood in tension, definitely not simple gravity load.
Armor the openings, strap the roof down to concrete, and you have a chance. But if you don’t, around 130-ish, any piercings add interior pressure to lift, or the tension fastenings holding down the foil shaped roof let go, and really, whether you keep your walls from then on is how much they suffer when the roof comes off, and random chance.
With those factors in play, there’s just not that much difference between 160 and 185.
I think we’ll se what we saw on Barbuda. Intact with maybe glass, shingles and a bit of roof deck gone, roof gone but walls standing, roof and longer or corner wall sections missing, and the odd bare foundation.
FWIW...the Barbuda aerials didn’t come close to the PMs 90-95% assessment. A very few areas, yes, but overall, eyeball squint put numbers at 25-30% major damage, less than 10% bare pads.
PS...no juice, people hurt and needing help, and any long distance radio antennas knocked out, even amatuer long wire types on poles or fences, usually add up to first word coming from outside intervention.
Casualties. They had less time to prepare. Mudslides. A few percent rate increases on each variable. Times 70 over what...3.5?
On the plus side. Maria didn’t really have time to pile up a solid Cat-5 surge.
I will never claim to be an engineer or physicist. However, Dominica topography would expect higher winds at higher elevations. Furthermore, as vegetation on hillsides is stripped, mudslides will impact structures with or without roofs.
This page is useful — certainly not the last word by any means, but simple enough that most people can grasp the math behind the sheer (not “shear”, for those wondering) forces involved:
http://www.wikihow.com/Calculate-Wind-Load
You are absolutely correct about openings, of course, and if the steel IS in tension, there had better be plenty of it. I am more familiar with tornado damage* - with hurricanes there is of course a much greater materials’ fatigue factor to consider.
*When I was a young man, our entire brick-face home was nearly pulled free from the foundation by what was rated a high end F3 tornado. (I personally think it may have hit low end F4 briefly.) As it was, most of our roof was gone, about 1/2 the walls were down, maybe 1/4 left severely leaning (say, 45 deg.), and another 1/4 were left more upright / vertical, but were “wavy” if sighted along horizontally. Even those last 1/4 of the walls were within generally about 1/2” of the big nails holding the walls to the foundation pulling free. A neighbor had their entire garage (which was separate from the house) lifted up off the concrete pad and dumped more or less intact into a pond in the rear of their property, with no scrape or gouge marks between the pad and the pond. Another (the owner was a quality homebuilder / contractor) had his attached garage sheared off their house and simply obliterated into small pieces. A 3rd had his car lifted up out of his garage (roof gone, walls intact): It somehow landed in the yard with minimal damage. Once debris was cleared, he drove it back into the roofless garage, and into town the following day. Freaking Wizard of Oz stuff I might not believe had I not seen it myself.