IMO the current bitcoin mining operation is centralized by the likes of Bitmain and some mining pools. As far as I have read, LN would lower the fee a lot, off-chain transactions would lead to more user privacy, and exchanges would not be prone to hackers. Payment channels/hubs getting centralized is a possibility, you would need a lot of revenue backup to operate hubs and some big players like Coinbase would dominate, but the same thing is happening now. Hard fork is not backwards compatible, and with the rate bitcoin is getting adopted increasing blocksize would be a temporary solution, what happens when limit is reached, another hard fork? Segwit alone is not a scaling solution, LN is. If it could be said that LN would lead to centralization then it is not wrong to say that bigger blocks would lead to centralization simply because only a few miners would be able to afford to run full nodes, centralized entities.
As a user, I do not see why not LN. I am not against bigger blocks, but LN could be the permanent scaling solution. Using temporary solutions and letting bitcoin users go through periods of uncertainty would be bad to the community as a whole. The prolonged scaling debate and increased fee has already affected bitcoin, alts have increased in value during this time, continuing this would be illogical.
LN does have a niche.. but if you actually look beyond the utopian propaganda illusion and see it for what it is. its not a solution for everyone. so dont be under the impression that LN is the solution... its just a OPTION.
as for the bigger blocks kills miners.. LOL
miners just use ASICS, which connect to pools.. an ASIC does NOT have a hard drive. an asic does not handle the blockchain, an asic does not validate transactions.
miners dont have any worries about the blockchain. all they get is a
few hundred bytes of data. that the asic re-hashes a few trillion times a second and then sends out a solution thats only a
few hundred bytes long.. not megabytes, not gigabytes..
as for nodes..
the fact that most dynamic proposals (if using real network consensus) would grow only when the majority of nodes accept such change.. then its not going to cause issues.
what is going to cause bigger node dilution issues is all the prunned/stripped/filter node crap that (using torrent analogy) means there is less seeds of full data for other nodes to leach from