Don't bother arguing with these fools as 99% of them are too lazy to have done any due diligence themselves and just hope that someone else will work on bitcoin and it will somehow make them rich.  From memory, when I did the estimates way back in the day, I think I came up with bitcoin needing 8 MB blocks in the long run to function as a global settlement system where transactions of around $10k or more are viable.

Is 2 MB blocks + LN equal or better to that?  No.  LN is an even more centralized, more permissioned version of bitcoin where they seek to transfer all consumer payments to.  So you theoretically CAN transfer most bitcoin traffic there, but LN doesn't actually have anything in common with bitcoin in the first place.  It's a totally different system.  It's like replacing bitcoin with bitshares and pretending it's the same thing.  But when it comes down to it, bitcoin transaction validators are designed to centralize, nullifying any reason for it's existence, so there's no point in even discussing building any 2nd tier layers on it.

Discussing LN vs bigger blocks is like discussing what color to paint the Titanic.  It's 100% impossible to create a decentralized digital currency where transaction validators don't centralize.

The current system where every user is a network node is not the intended configuration for large scale.  That would be like every Usenet user runs their own NNTP server.  The design supports letting users just be users.  The more burden it is to run a node, the fewer nodes there will be.  Those few nodes will be big server farms.  The rest will be client nodes that only do transactions and don't generate.