As you may know, Bitcoin (BTC) has a scaling issue. Transaction fees are up to around $10 these days. The problem is obviously known by the people who work on BTC, so what is being done about it? Well, it turns out not much, at least not on the base layer.
On the BTC network there is a "block" of transactions every 10 minutes (on average). These blocks are limited to 1 MB and there is absolutely no plan to change this. So the people that work on and want to scale BTC are 100% invested in second-layer solutions that keep the 1 MB block base layer unchanged. I will give two reasons why I seriously doubt the feasibility of second-layer solutions on BTC.
Off-Chain Transactions
For the first reason, the entire point of Bitcoin is to be able to store and send value to anyone without intermediaries or censorship and in an extremely secure manner due to digital signatures and the proof-of-work done by miners. It's basically a distributed time-stamp server that logs messages (transactions) and those messages can never be "unlogged". If you're careful with your private keys, as long as miners keep mining this decentralized server remains completely and utterly unhackable and immutable.
This technological breakthrough achieved by Satoshi Nakamoto 12 years ago is potentially the best form of money there has ever been. It's an extremely elegant solution to a problem that mankind has faced for millennia. It is possible humanity won't see another monetary quantum leap like this for hundreds or thousands of years.
Obviously there is great market demand for messages confirmed on the blockchain. Full blocks and fees up to $10 or more on BTC are evidence of that. For whatever reason, there is no plan to increase the throughput of the base layer of the time-stamp server, the BTC blockchain. So, by default, second layer solutions are the key to solving Bitcoin congestion and helping adoption take off.
Second-layer solutions are, by definition, *not* on the base layer. The goal is to perform as many BTC functions as possible outside of the blockchain itself.
Right now exchanges are by far the most widely-used second-layer solution. Few people would put it in such terms and that's probably not the underlying motivation for most or any exchanges, but they indeed comprise the most adopted second-layer solution. Inside of exchange ecosystems (Coinbase, Uphold, PayPal, etc.) users can do all sorts of stuff with their BTC without interacting with the blockchain itself. And when the time comes for the user to remove BTC from that ecosystem, there is a transaction that finalizes it on the blockchain. That means that the user can make all sorts of "transactions" in sort of a bubble, away from the blockchain, and the exchange keeps track of those "transactions" on their internal database. The result is that hundreds of thousands of BTC "transactions" can occur without involving the blockchain at all, which conserves space on the blockchain, keeping it mostly available for all the settlement transactions when people are done having a portion of their BTC in a particular exchange ecosystem.
The Lightning Network (LN) and solutions like it make up another sought-after second layer. The extreme basic idea is that it creates something like a tab between two people. Under optimal conditions, an arbitrary number of back-and-forth "transactions" can happen between two users outside of the blockchain. When they've reached some agreed-upon limit or one of them is drained of liquidity, then they can settle the end-result of their interactions on the blockchain. That simple tab between two people can be expanded upon to creates a series of interconnected routes of liquidity channels. The hope is that, with enough connections, any LN user can find a route to any other LN user. Then, depending on liquidity, any two users can make as many "transactions" as they want outside of the blockchain before making a real blockchain transaction to finalize everything. Again, this conserves the limited space on the blockchain so that people can use it when needed.
What is the common requirement for second-layer solutions? We can see that the goal is always to allow users to make an arbitrary number of "transactions" that are specifically not recorded on the blockchain until the parties want it finalized. What is the problem with this?
The problem is that such "transactions" before they are finalized, no matter how or where they are implemented, specifically do not and cannot benefit from the elegant, unhackable, irreversible Bitcoin time-stamp server that solves the original problem of storing and sending value without intermediaries or censorship. Such "transactions" are not part of Bitcoin and are not recognized by the Bitcoin network. Thus, they are hackable, reversible, and subject to intermediaries and censorship. They are therefore weak and susceptible to attack by malicious actors (from the public and private sectors) in various and unpredictable ways (and many predictable ways as well). "Transactions" outside of the blockchain completely bypass everything that makes Bitcoin beautiful. Obviously the goal is to eventually finalize the "transactions" on the blockchain, but before that final transaction message is broadcast to Bitcoin nodes and put into a block, they are not part of Bitcoin at all, but only pretend to be.
Inevitable Congestion
Now for the second reason I have for doubting second-layer solutions can solve Bitcoin congestion. 1 MB blocks can accommodate, at most, seven transactions per second (assuming 238-byte transactions). In an ideal BTC scaling world, the vast majority of those transactions would be settlement transactions. Whenever the demand for blockspace is higher than that, congestion sets in and fees start going up quite a bit (It's interesting what people consider a "high fee" these days. When I first got into BTC anything more than a few cents was considered absurd. Now a few of $1 or even more is considered acceptable by the BTC community).
Most of the time, the congestion and fees are not a HUGE problem if people are willing to wait for nights and weekends for the network usage to calm down (still completely unacceptable and unseemly from a UX standpoint, but we'll ignore that for now). But if demand ever consistently goes above seven transactions per second (even assuming 100% settlement transactions for exchanges and LN or whatever) and that demand is sustained over a long period of time, then there is technically no limit to how high fees can go. We're talking about a mempool that continually grows over time and never stops. Miner transaction prioritization is set up as a blind auction as no one knows what everyone else is going to bid for block inclusion with their transaction fees. In a continual high-demand environment (7+ transactions per second), those bids will go higher and higher and there is no upper limit besides the point at which users utterly refuse to use the network due to those high fees. Such a network is self-defeating and loses utility as demand increases.
Just to give an example to illustrate BTC's limitations, let us imagine that every adult in the United States (approximately 200 million people) wanted to use the Lightning Network. Even allowing for ideal conditions and only one on-chain transaction each year for each US adult (extremely unlikely), BTC's blockchain could theoretically handle that, but literally nothing more. Just blocks 100% filled with LN settlement transactions from each of those 200 million users each year, all day, every day. There would be no more room for any other transactions. Any more demand that that would quickly make the network almost completely unusable by all except the most wealthy.
If BTC cannot handle more than 200 million users (less than 3% of the world's population), even under completely ideal (i.e. unbelievable) scaling conditions, how can it handle worldwide adoption and usage? How can it facilitate a worldwide payment system? How can it change the world?
Unfortunately it cannot.
Reasonable Doubt
I believe these two reasons demonstrate why a reasonable person can and should be extremely suspicious of the traditionally-accepted BTC scaling plan. There is a better way and that way is actually scaling the blockchain for use by however many users desire to transact on the base layer.
A very nice picture of the broken coin.