I do not consider increasing blocks a good solution. It does not scale while the negative consequences for decentralisation are unclear. Since Bitcoins only true value lies in decentralisation, I support not hastily moving away from this core concept just to solve a current - perceived as immediate - issue. There is too much at stake.
Maybe you are right, bigger blocks do not scale. But increasing the current size is a requirement for climbing. Check the LN white paper and note the block size needed so that each person can open and close a channel once a year using segwit.
Yes, blocksize will eventually need to be increased. I hope it will be done right after LN is running smoothly and widely adopted.
But... just a small/moderate blocksize increase... I cringe when I hear about those "lab" tests carried out by Bcashers with up to 1GB blocks!!!
First deploy the exponential scaling solution (LN). Then adjust the blocksize (linear scaling) as needed for having a non-congested backbone.
I can admit that, if this was an enterprise, the blocksize increase would have come first as a temporary fix whilst the real scaling solution (LN) was being finishing development/tests/deployment/adoption etc... But it is not... decentralization has its pros and cons.