Bitcoin stands on three pillars, the users, the miners and the developers. The weakest link here was the developers but fortunately Bitcoin has self correcting features which resulted in the forking off to a new development team (Bitcoin Cash).
Lightning destroys Bitcoin's security model.
One of the most fascinating attribute of bitcoin was that the system pays for its' own security. It may seem trivial now but ask yourself back then, how do you pay someone to secure your protocol with coins that are worth nothing? This was the chicken and egg problem. All prior solutions, including governments, have all relied on trusted entities to perform that function.
Miner's reward = Block Reward + Transaction Fees = 12.5 + E[xT]
The principle idea is that as the Block rewards decreases every four years the block reward will be replaced by transaction fees.
T is allowed to increase as fast as the technology for block space allow while keeping x as low as possible.
This requires larger block sizes. Even one that can accommodate billions of transactions (T) in say 10 years. We can be sure that technology will come to the rescue. For example, as we will move to 5G networks, developing countries will skip 1G, 2G, 3G and even 4G to move directly into 5G with the rest of us. This development will happen as sure as they skip landlines entirely to move into mobile networks.
Lightning network changes the security model to :
Miner's reward = Block Reward + ( Transaction Fees - Lightning Fees ) = 12.5 + { E(xT) - E(yL) }
As the block rewards decreases the transaction fees must also increase. The idea here is to make on-chain transaction very expensive and move all smaller transactions to the Lightning network or side chains.
T is capped and x is allowed to increase as much as demand for settlement allows.
L is allowed to increase indefinitely and y is kept very small.
Transactions to get on and off the lightning channels are part of total T. If it is not obvious yet, as x increases it also gets more expensive to open and close a lightning channel. So how would one get on to lighting if it gets too expensive to open a channel. You will have to subscribe to a channel node directly. That is put your money into a "banking" node. These "banking" nodes will now be in a position to allow or deny services to you ie Be A bank!
At the moment we tend to assume that miner's cost are denominated in dollars or fiat. Therefore as the transaction cost grows to thousands of dollars it will keep pace with costs and miners will be happy. But will they? When we get to the situation where the unit of transaction is bitcoin and not dollars, then the block reward will tend to zero in BTC terms. The only real reward are the on chain transaction fees.
Miners are price takers. They get to determine their income by selecting from a smorgasbord of fees. "Banking" nodes become price setters. They put up that smorgasbord of fees. It will be in their interest to set Lightning fees as high as possible and Mining fees as low as possible.
Nodes were never meant to be compensated in the Bitcoin protocol. They are placed on the user side of Bitcoin's 3 legged equation. They pay for the cost of running a node because their business model requires it. It is their cost to use the system.
Lightning has not taken hold yet and we are far away from "banking" nodes. It could work if miners do not have a choice. But miners have a choice in Bitcoin Cash. Miners will not put up with a situation where they get elbowed out of the system, just as happened with the original core developers.
Bitcoin's security is one of its' greatest strength.
Bitcoin uses proof of work to secure the protocol. This is the most secure system we know. Nothing has ever worked before bitcoin. To emphasise this point, the fastest and most powerful chips are used to mine bitcoins even before they are deployed into mainstream computers. The technology to secure Bitcoin is bleeding edge and will remain so.
Any departure from this is a compromise and introduces some level of centralisation or trust. Proof of stake requires some form of centralisation, moderated in different iterations of the POS model. Without Bitcoin coming first, POS cannot take off on its' own. Bitcoin made it possible for value to flow into the POS tokens.
Trust but verify
It can be argued that not everyone can run a Bitcoin Cash node because of the cost and we have to trust a small number of these expensive nodes. However, any business or organisation that needs to verify their transactions real time will have to run a node. We could have hundreds of thousands of nodes across all sectors and industries. We don't have 100% trusted nodes but we have enough, just as we may not have 100% honest miners, but we have enough.
We cannot say the same for "banking" nodes because their profitability depends on their control of miners revenues, and thus the erosion of Bitcoin's security model. You will not be allowed to verify or audit their setups.