This is part of why Solana has been making such inroads the past couple of cycles.

Shopify and others are now using SolanaPay to do USDC transactions and they settle in under a second for under a cent. It takes a very beefy server to run a Solana validator, though, so many worry that it’s insufficiently decentralized.

I suspect that in the longer run, most of the value will be stored on Ethereum and compatible chains while most transactions will be done on Solana and newer chains.

payments is a meme for every bagholder coping about their cheap blockspace chain. literally every cycle cheap l1's come up

eth mainnet is lindy as the most secure settlement/data availability layer, other layers for execution settling back to mainnet or other da layers potentially solve settlement fee/l1 data writing rent issues

Ethereum L2s are just as fast and in a few months will be just as cheap.

Perhaps the “store of value” use case. Buying and selling physical gold bars adds significant overhead costs as well.

But generally I agree with you, the fees are a big problem.

Eh, ransomware payments would tolerate these fees

L1 (Ethereum Mainnet) fees spike that high, L1 is the highest secure layer of the network and inherently higher price during peak usage. Pick a flavour of L2 of your choice and enjoy cents per transaction. See prices here https://l2fees.info/

L2s are definitely the way. They publish all state, state updates, and proofs of the validity of the state updates on L1. That means they publish all consensus critical data on L1, allowing them to be as secure as any L1 smart contract.

They're not yet production ready though. All use centralized coordinators and failsafes, as a precaution against catastrophic flaws in the smart contracts they have deployed on L1, that could lead billions of dollars worth of digital assets being stolen/lost.

Arbitrage and speculation seem to be the usecases that survive (and thrive) under such conditions.

I find those to be worthless usecases to society in general, but they are usecases.