Using eltoo along with a SIGHASH_ANYPREVOUT Bitcoin upgrade and channel factories this is certainly doable. At this point you are not dealing with lightning penalties anymore nor are you concerned with the overhead of previous state transactions that lightning nodes currently have to manage and store. In addition, channel factories with 20 users is likely a max you would want to deal with due to issues of people going offline. However, channel factories allow channel re-balancing and allow people to open multiple channels at once, increasing their connectivity. A 20 person channel factory opening via a single taproot address allows 100 channels to be opened. Yes, people going offline is problematic when you consider the fact that people in channel factories have channels opened with multiple people, then waiting a day or two for the other person to come back online is feasible. In addition, with channel factories you can splice out the non-responsive person and continue with a new hook transaction.
As for onboarding, you can onboard 6 billion people in 90 days with just 20 person channel factories. If the running narrative is "Get the world onboard as quickly as possible" then protocols can work with exchanges to achieve this with the least transactions possible. Consider 20 people using an wallet app to connect and form a channel factory. They could very easily work with an exchange like Coinbase to get their fiat converted into BTC. The 20 people would combine their 20 public keys into the 20 of 20 multisig hook transaction address and give it to coinbase so they can send the sum of all their funds to the new address. This would effectively onboard 20 brand new people onto the Bitcoin/Lightning network at once via factories.
Given that hundreds of thousands of people would be doing this with the exchanges since "everyone wants to get on bitcoin", the exchanges would be incentivized to minimize their fees by constructing large transactions. They could construct a 1 to 23,250 P2TR transaction that was the size of an entire block. With the sum of fees from all the participants, the exchanges massive transaction fee would be more than enough incentive for the miners to mine it.
Given than each of the 23,250 outputs represents 20 people via a 20of20 multisig channel factory hook transaction, then that would mean a single Bitcoin block could onboard 465,000 brand new people onto the network. At this rate you would be able to onboard 6 billion people in 90 days. (Note: I calculated the max transaction output of 23,250 using the bitcoin core max weight size of a block at 3996000WU)