It's a dead-simple idea. You increase the supply side of the fee market and users don't have to compete for transaction finality anymore. It has been implemented and we know it works.
Of course, that is only kicking the can down the road and has many caveats. It still has proven effective to complete its goal (minimize tx fees).
Lightning networks is a WIP that is months, perhaps years, away from hitting production. As such, I wouldn't call them the "main approach" to scaling except in the mind of core engineers. It might drive tx costs down, it might not, at this stage it is too early to tell for sure.
As such, I think OP is right. It is completely fair to say that block size increase is the "main" approach, as in the "the one that we know for sure works right now", to making Bitcoin usable as a payment system.
Now on a side note: I don't think you needed to be that aggressive with OP. I won't address the attacks on Roger Ver because it is my belief that attempts at turning a technical debate into politics should be met with contempt.