Simply put: it isn't worth it.

So you're about to risk ruining a half-trillion project (at the time or writing this), opening up the Pandora box of a debate that broke Bitcoin in half already once (and in multiple small pieces), completely ignoring the mountainous history of second layers, breaking backwards-compatibility, prioritizing the short-term in a project whose success comes from long-term conservatism... And all that to... Quadruple the on-chain transactions? What when that will neither be enough? Will you start over?

Full nodes operators needing an extra terabyte disk is the last concern of mine if such thing was ever going to happen.