I would say UX friction comes second after adoption. The lack of adoption is really the biggest reason. Lower fees are all nice and good, even if you have a decent UX design that allows even newbies in crypto to easily navigate their payment experience but if no shops are offering Bitcoin payments then lightning is not going to be any good in the first place.
Lightning is still too early, basically.
But it has the chance to grow, adapt and become more amazing and will be ready for the future adoption.
Lightning UX on mobile devices is great. Try to onboard your local vendors, you will see that it is pretty easy.
Real-world peak bandwidth: 98% saturation at 1.5 Mbps upload (per Core v25.0 node logs). Why home nodes choke: Even 100 Mbps connections max out during mempool spikes (e.g. July ETF rush). Dedicated servers? Smooth. But 67% of public nodes run on home broadband (mempool.space 2025 data, hope it is accurate). Tech fix? BIP-325 (compact blocks v3) cuts bandwidth by 40% - live in Core v26.0 (Q4 2025). Let’s build a BTC Scaling Pulse report: I’ll draft metrics (node bandwidth, mempool depth, orphan rates and so on ) if you co-author. First edition by Oct 1?
This has nothing to do with blocks though. If we were to go with that reasoning, then there is never going to be a technological justification for increasing the block size. At any bandwidth someone could spam the mempool so much to choke it.