I noticed a similar phenomenon monitoring Matthew last year. The winds aloft, that are taken to be the true measure of the storm, aren't at all the winds that reach to the ground.
I've noticed this in multiple hurricanes over the years. Not sure how the NHC calculates the differences between flight level and surface winds but it appears their ratio is wrong -OR- combined with the 8 min average vs the 1 min average the numbers are radically different.
I've always though the NHC should adjust the "estimated" winds down to true wind speed once they had good ground based data (buoys or other known calibrated weather stations). However they seem to stick with flight level winds. I do believe some of this is CYA, better to estimate high then to give people a false sense of security with a Cat lower forecast. In the past I've seen report of gusts, micro bursts, mini-tornados, down drafts and other small-scale higher intensity winds in certain areas due localized environmental conditions during land falling 'canes.
The difference between a 1 minute average and an 8 or 10 minute average is actually pretty dramatic. I'm not sure which one would be most 'representative' of the storms potential damage.
The issue with surface winds as a ratio of winds aloft is because of turbulence and how well the wind 'mixes' down towards the surface. Also of note, the 'surface isn't defined as the actual surface (but instead 2 or 10 meters above the actual surface). which is why there is often a difference in the wind speeds at the theoretical surface versus the actual surface.
Trying to model actual surface condition (as opposed to 2 or 10 meters up) is incredibly difficult, you have all sorts of microscale phenomena to deal with, such as frictional forces and turbulence from objects, which makes the modeling that much more problematic.
As far as mixing from the "upper levels" (850mb and lower pressures), There are lots of papers on TC wind mixing exploring that very issue. Since TC winds are not well mixed, some of the questions that is still being answered are: how do the winds aloft mix down to the surface and when can the mixing ratios of winds aloft be applied to determine surface winds?
(At least, I don't think there are definitive answers on those...)
So in short, they don't actually simply stick with flight level winds, They use surface obs when possible, only resorting to ratio estimates when there isn't any other option. And because of the nature of the winds, that there is a lot of variability in wind speeds even in locations that are relatively close to each other spatially. which can make a wind forecast look bad even though it's representative of the maximum wind speeds in that area instead of every single location in that area.
M. S. Earth and Atmospheric Sciences, Georgia Tech - May 2018.