The loud voices tend to be most wrong when they talk about things that aren't Bitcoin. Which seems to be a bear market engagement farming phenomenon, so fairly harmless.
I hope we see more and more robust lightning tooling. Even in the two years or so I've been using it, it's come on so far. So in five I'm hopeful that the onboarding and channel management / liquidity issues (for newbies looking to spend as well as merchants looking to take payment) get smoother. Spliced channels and LSPs are only getting started.
Would also love to see more privacy progress, possibly some convergence between lightning UX improvements and Cashu maturing. The pieces are all in place to do without shitcoins like Monero.
Projects like SeedSigner are also exciting and I'm hopeful they can keep funding and support.
My wild dream would be for some serious work to go into making Bitcoin Core a viable newbie desktop tool. The eternal sticking point here is that it doesn't support seeds, for thoughtful reasons, but it's a big barrier and means newbies have to manage both Core software and another coordinator. Perhaps rather like the HWI support in Core, there could be some kind of mnemonic interface without compromising the system?
I'd like to see the end of people thinking they need to use raspberry pis and node kits, which actually only complicate and increase risk.
One thing I worry about, particularly for lightning, is how American centric the viewpoints are. For most of the world, for example, NFC payments, free instant bank transfers, and low cost debit transactions are already a reality. So understanding that lightning has to go far beyond feature parity.