You mentioned in the other thread a couple of points which, personally, I find to be "reaching" when it comes to actually identifying what I would consider to be "real" issues with LN.
For instance:
* Very difficult to use and explain to users. No way your mother let alone grandmother is going to understand all this.
That really isn't a valid argument... I'm fairly sure that at least 90% of people who use the internet every single day have ZERO knowledge of TCP/IP, HTTP(S), SSL and other technologies underpinning the internet... and yet they're able to post selfies and pictures of their cats/grandchildren without too much issue.

Likewise, I doubt that many people understand the inner workings of internal combustion engines or how a clutch/transmission works... and yet they drive cars around.
The challenge there is not that people need to understand the underlying tech... we just need to provide applications that enable them to use it.
* You still have huge fees if you want to wire money into/outside the channel.
* It's also not feasible for very small payments/channels. If you open a $30 channel with your coffeeshop to buy a few cups of coffee per week, then the price of your coffee doubles because of the huge fees to open/close the channel. Your only option is to route and HOPE there IS a route.
You mention "huge fees"... exactly where are these "huge fees"... last time I looked, there are transactions getting confirmed with 2 sats/byte fees... OMG... outrageous! /s
Sure, fees have been ridiculous in the past... but that was because the mempool was flooded and the whole supply/demand thing kicks off. Now imagine that the vast majority of those transactions are no longer in the mempool and are shifted "off chain" onto the LN... what is likely to happen supply/demand-wise then?
Less on-chain transactions = less "demand"/same "supply" = lower fees... That's kind of the whole point of LN.