The lighting network is bound to fail because it suffers from the same problem Bitcoin is solving, i.e. it only works in centralized hubs. Think about how a new user would come to Bitcoin if the LN was in full effect, they would need to spend hundreds to open a channel to a centralized hub, the more centralized the better so it can connect to more places, so they can ensure they have the funds and the connectivity to use the network, because channels can't be increased, if he ever needs more funds than what he has he will need to close the channel and open a new one, and because channels might be used by third parties when routing through the network if he spends double that amount to create two channels to two centralized hubs he risks having their funds go from one channel to the other and having to go the long route for his own transactions.
Also the idea that running a node is more expensive because of larger blocks is mostly nonsense, for starters the only nodes that matter are mining nodes, even if your validating node finds an issue it has no power to do anything about it. Secondly there is such a thing as pruning old blocks to reclaim space. Third the whole point of Bitcoin is that you don't need to validate transactions because mining nodes have incentive to stay honest, and the regulators are other mining nodes who stand to gain from others dishonesty by mining the same block they did, and again, mining nodes require lots of hardware, some extra HDD for old data is cheap in comparison. Fourth, why don't you feel the need to validate the LN? Why just validate the on-chain stuff but trust the LN? Surely you'll want your validator Node to process all of the LN transactions to ensure they're valid, no? Or do you trust that the LN works? And if you do why not trust that the technology the LN is built on top of also works?. Not to mention that even in poor countries the cost of a hard drive that can hold years of data is lower than a couple of transactions on the main Bitcoin network during peak times, and a lot cheaper than opening one channel on the LN that's worth using, for example earlier this year the average transaction fee peaked at $123 https://ycharts.com/indicators/bitcoin_average_transaction_fee so of someone wanted to make a transaction that day on the Bitcoin network they would need to spend $123 extra, that is enough to get a used 8TB HDD which should hold all of the BCH Blockchain just with the cost of a single transaction.
Long story short:
LN has design flaws and doesn't work
Even if LN worked, on-chain transactions are better, less limited and more secure
Even if those design flaws could be addressed, the main chain will still become unusable due to high cost of entry unless blocks get increased
If blocks never get increased, popularity will be the death of Bitcoin like it happened in 2017, if someone needs to throw away hundreds of dollars just to get into the network, they'll never do it. And because RBF people that don't pay whatever the current tx fee is can get scammed.
Validator nodes serve no purpose
Even if they did running a validator Node on BCH is cheaper than using BTC
Even if validator nodes made some difference then the LN would also need one and then the amount of transactions there would increase the need of a node to higher than what would be needed for BCH (since besides regular transactions you would also need to validate on and off channel transactions).