What was concluded in that article from 2011 was to just repeat the conclusion from the 2004 article which was:
So, GPS and military communications would be unaffected as would our spy satellites. A few decades ago a lot of telecom companies used satellite in their networks, but that's almost all gone to fiber now except for extremely rural areas like parts of AK and remote portions of the SW. The same is true for a great deal of video distribution, although I suppose we'd lose Direct Broadcast Satellite video (i.e. dish). Even things like most vehicle tracking have gone from satellite to cell networks. There is a bunch of telemetry data for remote oil wells sent over satellite, but that's hardly critical.
Which LEO satellites are you thinking are critical infrastructure?
Well, first of all LEO isn't a specific altitude. It's an altitude band, 99 to 1,243 miles. That's quite a range of altitude.
Medium Earth Orbit is 1,234 to 22,236 miles. And geostationary is 22,236 miles. High Earth Orbit is above 22,236 miles. (For those who don't know...and I had to use google-fu to refresh myself on the specific altitudes, myself.)
Earth observation satellites, including spy satellites, use LEO for resolution purposes. It's much easier to get high resolution images from LEO than from higher orbits.
Many communications satellites use LEO. LEO means lower power requirements, not to mention faster communications times. Some do use geostationary orbits, which are most definitely NOT LEO. But all your communications via satellite constellations are from LEO.
Polar Orbiting meteorological satellites and the military's DMSP meteorological satellites orbit in LEO.
And let's not forget that those occupying the ISS in LEO probably wouldn't appreciate this, either.
How much of the satellite infrastructure in LEO would be affected would depend largely on the size of the nuclear detonation and the specific altitude in which it occurred.
How badly would the loss of 90% of the LEO satellites affect us planet side? In one aspect, you are correct. Their loss wouldn't likely affect critical infrastructure much. A lot of short-term nuisance, I'm sure, but a very costly one nonetheless.
The other beauty of LEO is that it's cheaper to get replacement satellites back up there. So it's not as if any damage wouldn't be taken care of in relatively short order. But it's still a rather large loss to the organizations which use and maintain the satellites in LEO.