No. Propagation of data. And the equation of orphan rate applies. And different miners see a different orphan rate. This is what causes centralization and impacts scaling.
"centralisation"??
hmm 2013.. blocks 0.5mb ... 5 pools
2017 blocks 0.999mb... 20+pools
goodluck in your research
p.s
i do like how you now stepped aside the PoW statement.
and are now trying to meander it to being about tx validation.
EG
if PoW algo was to change. and only PoW was to change
if it took 2 sconds to validate tx data before
it will take 2 seconds to validate tx data after.
the PoW is not about tx data. is about the block hash's security lock that comes AFTER tx validation.
and that security lock is always going to be based on a 256bit piece of data no matter how many tx's are included or not.
anyway to gt to the OP's opinion
LN has a niche for some people that do spend more than once a week, such as faucet raiders, gamblers etc. but LN requires a deposit and eventual withdrawal (2 onchain tx's) by which due to uncertainty of how often some people wish to spend bitcoin would not want to waste 2 tx's to lock funds into a contract they may not use for that fortnight.
yes LN can have a 1 year contract. but then there is the issue of. can people predict their spending habits of a year to know how much to put in at the start. can they really trust having their whole hoard locked in for a year.
there are actual issues with LN.
such as address-reuse signing and blackmail, CSV changebacks etc. which is where even LN devs have suggested only use $60 amounts with ~2week contracts.