I would rather he tested floor standing speakers, but his conclusion (favoring absorbing isolation, not hard transfer spikes) matches my own conclusion after many years.
Years ago, I thought I was so smart when I removed the wheels of my JSE Model II’s and installed spikes (on my springy wood floor).
SOLID, all intended movement gets to the cones, right? All that weight onto 4 small discs is soooo much more weight per area, just tons of force, right?
After a few years, I put 3 wheels back on, (two front, 1 center rear, with rear anti-tip corner blocks just shorter than the wheel, only contact if the speaker tilts while moving it). I didn’t measure with a mic, but I heard no detriment, and got back the ease of positioning and alternate toe-in.
https://www.audiogon.com/systems/11516
BTW, dual wheel furniture casters, the axels do not wobble like single wheel casters do (even high quality ones I tried). The block above the 2 front wheels tilts the speakers back, aiming the tweeters to seated ear height.
In this case, the skirt acts as the anti-tipping blocks in the rear corners.
So, my 6 wheels are hard plastic, the contact area small, not absorbent, it's a matter of enough weight relative to the wheel's diameter, so they stay put, but you can move them readily.
I gave the spikes to my friend, he has them on floor standers, on a concrete basement floor, no movement possible, and his imaging is great, only if you sit in the middle, I would put his on 3 dual wheel casters like mine, slightly smaller diameter for his, to use an alternate toe in for 2 listeners. His are narrow, taller, tweeter at seated ear height, so no tilt is needed. I prefer shorter, both inward angle and tilt to alter the angles of reflections off the floor, ceiling, side walls, especially minimize woofer floor reflection