The security tradeoffs (segwit/lightning network) are not worth it to me. And as mentioned before, there's that entire lack of a working routing mechanism thing.
The fact there's been somewhat of a conspiracy to pretend lightning network can function in a decentralized manner will put bitcoin in it's grave once people figure out it's not possible. You would need to put all transactions in the same common que to prevent attacks and other problems, then you're right back to square one of on-chain bitcoin scaling. But the fundamentals of on-chain scaling are also bad.
You would need around 8 MB blocks for bitcoin to have market penetration into the upper middle class of the world, and even with blocks that big, only enormous transactions of something like $5-10k would be economical on-chain. People would hold, say $100k in savings in an address they control and transfer $10k at a time onto a 3rd party bitcoin debit card.
The problem with this solution is that bitcoin would be functioning solely as a settlement layer and it cannot compete with gold and silver as one. People would be far better served transferring $10k worth of metals at a time onto a 3rd party debit card instead. Metals are far superior as a store of value and bitcoin also has built in middle men (transaction validators) and doesn't remove counter party risk. Bitcoin's fundamentals as a settlement layer are essentially zero due to better options that already exist.