There are services like strike that offer custodial lightning channels, and will aggregate and send onchain tx's: https://medium.com/@ACINQ/introducing-strike-a-stripe-like-a...
While the network has such low fees as today, there is no need to increase the blocksize. Increasing the blocksize (or time) effects scalability in epic ways -- ie: Ethereum requires at least 500GB SSD to run -- so the minimum cost of entry is at least $300. Meanwhile 10m/1MB blocks still runs on $20 RPis.
I am not against mods to consensus to increase the block size, but there are more critical problems like solving tx malleability and new tx format (segwit) that provide more value today.
Dont be an ignorant big blocker. Look at the smartest people in this space. The real technical folk. What are they saying?