This is a great comment and I really appreciate you taking the time to write out a very clear description of the technology that RaiBlocks runs on.

I noticed that one of Crypto Investor's points was that scalability is something that is always solved in time, and though it is somewhat true for centralized technologies, the case for Bitcoin is far from the typical, and the LN is not even close to being the decentralized scalability solution that Bitcoin has been searching for.

First of all, how does LN work? If Alice wants to send money to Bob, they have to open up a payment channel on LN, which involves both of them sending money into a special Bitcoin address, after which they can exchange bitcoin so long as the channel is open, and eventually close it in order to "commit" the transactions on the main blockchain (hence the "off-chain" nature of the LN).

You can link multiple channels like so: if Alice <-> Bob and Bob <-> Stacy, in order for Alice to pay Stacy, Bob would have to pay Stacy first, and Alice would reimburse Bob (in order to see why this actually works in practice I recommend reading the LN whitepaper or this series of articles on the LN: https://bitcoinmagazine.com/articles/understanding-the-lightning-network-part-building-a-bidirectional-payment-channel-1464710791/).

All of this is supposed to be over a distributed network, but unfortunately LN will never be in a position to scale to a million users, let alone 10/100 million.

First: in order to create a payment channel, you actually need to do an on-chain transaction in order to place bitcoins in escrow. If you plan on only doing off-chain transactions, it's basically not possible, because these bitcoins must be sent on-chain in order for the channel to get set up.

Second: payment networks don't scale. Suppose we want to support 100,000 users. In this iteration of the LN, we can have a user with one payment channel to 10 other users, with no redundant connections (i.e, no cycles in the network). This means that the maximum amount of "links" we need to make in order to set up a channel between two users U_1 and U_2 is 5, but we needed 10 payment channels to have this property.

However, this isn't the only setup for the network. You can drop the number of channels from user to user, but you'll have to increase the number of hops - for example, if instead of 10 users we connected each user to 2 other users, we'd have a binary tree which would have depth log_2(100000) ~= 17, meaning that we would need at most 17 hops in order to connect any two users in the network with a payment channel.

A large number of channels (in CS terminology, the branching factor) means that users would have to divide up their funds over all of these channels in order to even use them. A large number of hops means that all of the users' bitcoins will be locked up routing everybody else's bitcoin. Especially since, in a hop, there is effectively a loan going on because we use network time to eliminate risk: Alice can only send money to Bob once she is sure that Bob has paid Stacy, otherwise we'd be in trouble.

In addition to this, every single hop that we make over the network must have sufficient funds in order to make this lend possible - for example, if Alice wanted to send 1 BTC to Stacy but Bob doesn't have that much, we can't use Bob for this payment channel, and have to find another one.

So you can see that, at least from this very quick explanation, the LN makes nothing better for (decentralized) scalability. It lends itself very well to centralization because then exchanges can act as the middlemen that ensure that enough bitcoin is in all the required payment channels so that merchants are able to use them, but I don't see any merchant ever using the LN anyways, because it has to commit some amount of Bitcoin for each channel, which is not feasible for many day-to-day businesses or even big corporations (Starbucks, for example).

So my point is, if RaiBlocks has indeed "solved" the scalability problem, it will be extremely good for crypto as a whole because we can actually use the damn thing to pay for our groceries.