Forgot to answer here, sorry

Thanks, but afaik this only shows that "there are hubs in LN" and that the network would be less efficient without them. It doesn't show that there is a general tendency towards bigger hubs and that's what I wrote.
But this does not make it centralized. It's quite obvious that only a few participants will have nodes open with more than, say 2 to 4 other nodes. For most participants even one channel is OK.
That it would be a "perfect utopia" like some kind of social network was never realistic, this would also make skyrocket onboarding costs.

I agree actually with you that LN is not a perfect utopia. It's a tool, but a potentially very useful one. I consider it a decentralized prepaid system to bundle dozens, in some cases hundreds of smaller transactions (not millions!) into one.

This is also not showing centralization. It is simply in a stagnant, very (!) slightly retroceding state, where existing hubs may minimally increase their importance. Newest numbers hint to a stabilization. But the current adoption is also not critical because on-chain fees are still palatable. I would have expected a growth however last year during the Ordinals wave, but yeah, we may not be there yet because the tools are still too difficult to use.

By the way, that image about centralization/decentralization your shared - that's exactly what our beloved big blockers were sharing for 7 years now.

That's what I wanted to add still to this thread, then for me also everything has been said, at least regarding these few issues.