If a user gets rekt using a layer 2 blockchain, he's 100% responsible and accountable for his losses and cannot blame anyone else as these protocols are mostly referenced as “off-chain solutions” for a reason.
They handle transactions off the mainnet, on a secondary chain, a sidechain if you want to get specific but these transactions are still termed “off-chain transactions” because they only seek finality on the Ethereum mainnet(for EVM L2s) at a later time.
That said, layer 2 blockchains also centralize liquidity because enjoying cheaper fees on a secondary protocol involves moving liquidity(your individually controlled assets) from the mainnet to a single smart contract account serving as a bridge account or channel and often controlled by a few people as a multisig wallet, essentially making themselves easy targets for attacks and exploits, and they are, as a protocol, centralized.
They are not decentralized blockchains and should not be expected to be one because that would just be dumb.
They are, by not just definition but by design, just centralized protocols. If layer 2 blockchains were capable of being decentralized or if they ever get there, they will literally just cease to be an L2 because they would have reach a point of being high performant chains that can and should be stand-alone master chains, avoiding being tied to some other broken and/slow patent networks, like Ethereum.
Why settle for being a layer 2 blockchain if you pretty much are everything the L1 is supposed to be — an immutable, permissionless and decentralized settlement layer — whilst also being cost-effective and scalable?
Why bother developing a literal competitor chain that is the better performer and settle to being a sidechain to a slow L1?
You see where this is going?
At any point that an L2 becomes decentralized, all would move to be a master chain and only maintain neutral bridges to compatible L1s and sidechains.
L1s are band-aids to deep wounds
I have always considered layer 2 blockchains to be a temporary fix to problems such as high fees and scalability, even when speaking on the lightning network, in the case of bitcoin.
These protocols are just band-aids to deep wounds that should be solved in the parent layer.
My argument has always been that a significant growth in layer 2 blockchains would be a direct threat to the L1s it was meant to scale. This is because L2s are not really scaling layers or scaling solutions as they are marketed, they are quite simply just a channel to reroute liquidity through, saving time and costs but in the process, stiffening the L1 market, leading to a reduction in users and revenue, because the action now happens off-chain.
Think about it for a second, how really does any EVM-based L2 scale Ethereum?
The problem, I believe, is in the absence of real education on what scaling a decentralized blockchain really should be about, because most people really don't know what “scaling means” so I'll leave a definition, courtesy of our AI friend, ChatGPT.
"Scaling" refers to the process of increasing the capacity, efficiency, or reach of a system, technology, or process to handle greater demand without a proportional increase in resources or costs. It ensures that something can grow effectively while maintaining or improving performance.
From reading this alone, one should be able to figure out that layer 2 blockchains don't scale the master chain in any meaningful way, but to ensure that it's really clear, let's look at AI and machine learning platforms.
Everyone understands that these systems are resource intensive as their models require vast amounts of computing power, storage, and real-time data processing.
So what are secondary solutions that help scale these platforms?
—Cloud computing services like AWS, Microsoft Azure and Google Cloud.
Before cloud computing, organizations had to rely on on-premises infrastructure, which had several limitations including high costs to set up, limited scalability and geographic constraints, amongst others.
Cloud computing solves these issues by offering on-demand, scalable infrastructure for AI and ML applications.
Note, at no point in this process does these scaling solutions steal users(customers) and revenue, they quite simply just offer solutions that enable these systems to best serve their customers for a fraction of the cost of building from the ground up.
Why is L2s different?
Layer 2 blockchains are different because their design isn't even about scaling the master chain, it's about giving it a break by simply taking away from it.
Apps on Ethereum for example don't automatically work on its numerous L2s frictionlessly so how would L2s be considered a scaling layer with any meaningful value-add?
DApps developers have to manually deploy on these L2s and users have to manually bridge and switch chains. All of these processes involve moving to a new platform entirely, which will feel foreign in one way or the other, even if they are EVM-compatible chains.
Settling transactions on ETH means nothing, they can pretty much settle anywhere they like. It's a piece of software that can be modified. L2s are not scaling solutions for some master chains, they are just stand-alone protocols stealing liquidity from those chains and hurting their revenue(essentially economy) in the process.
The only difference a user should feel when a system is being scaled should be the increased performance and lowered cost, if applicable, not having to deal with bridges and hundreds of billion-supply VC-rug tokens.