Any centralization or decentralization of the LN is not relevant as some sort of showstopper IMO.

As others have alluded to before in previous threads and essays, there may end up being two sorts of networks - one focused on privacy, more decentralized, and with more nodes - and another with more centralizing factors, higher capacity, and adhering more to things like AML, KYC. I don't see any issue with this, and as everything can be custodial, anyone can enter or leave the network as they see fit.

The technical scaling Lightning provides is required for any sort of hyper bitcoinization from a medium of exchange perspective. Either the largest exchanges become the Visa's of the world and keep a ledger among their businesses and settle on some sort of regular interval (giving them a position to rehypothecate and in a sense "pay" for the services to maintain infrastructure), or the LN grows to support and chunk away at the functionality Visa provides - albeit from the ground up in many emerging markets where having a new financial system may be more of a net new opportunity and provide utility.

But for Bitcoin to become a currency beyond a digital store of value, there will need to be some sort of off chain immediate exchange property, and the success of LN helps ensure we don't face rehypothecation issues.

The biggest challenge I see in the near term are demand factors - getting more goods and services denominated in bitcoin (and LN), which in turn gives more incentive to routers, which encourages network growth, and makes running a node more profitable - especially considering the time, expertise, risk, and captial node operators are putting forth in these early days.

This is why things like Taro are so crucial and helpful for network growth and adoption as well. It's a necessary demand flywheel to keep the network effects headed in the right direction for everyone.