Why is it absolutely impossible for DA to approach 50% of L2 costs?

This is a very good analysis and also the most reasonable bullish reason for DA. But I think DA can never be close to 50% of L2 costs.

Due to structural economic reasons, sorting always generates greater value than DA......

Block Chain is basically a business that sells Block space. Due to the difficulty of exchanging Block space between chains, they have formed a near-monopoly.

But not all monopolies can earn supernormal profits. The key lies in the ability of consumers to differentiate prices.

If there is no price differentiation, monopoly profits are unlikely to be higher than commodity profits.

Think about how airlines differentiate between price-insensitive business travelers and bargain-seeking consumers. Or how the same SUV model is sold under vastly different prices with the brands of Volkswagen, Audi, and Lamborghini.

Priority fees are an incredible price discrimination mechanism in blockchain. The fees paid for the highest priority transactions are actually several orders of magnitude higher than the median.

Both L2s and Solana achieve high throughput and high income by using transaction priority as a price differentiation.

The cost of marginal trading payment is very low, achieving huge TPS. However, insensitive trades will receive commissions and pay most of the network income.

The following is the distribution of 5 Blocks randomly selected from Base L2. This is a clear Pareto distribution, which makes price differentiation very effective.

The top 10% of transactions account for 30% of revenue. The bottom 10% of transactions account for less than 1%.

The problem is, although the sorter made a lot of money, the DA layer did not participate because it has no price differentiation ability.

That high-value Arbitrage for ETH is charged the same as 1 wei junk email transaction because they are settled in the same batch.

Due to the very low value of marginal trading, you can only achieve high TPS when the intermediate transactions can be on-chain at close to zero cost. However, with DA, the payment for each transaction is basically the same.

The DA layer can have high throughput or high revenue, but the two cannot be both.

This makes it basically impossible to scale rollup without the collapse of ETH network revenue.

The rollup-centric roadmap is fundamentally flawed as it abandons the valuable part of the network (order) and assumes it will return to the worthless part (DA).

I was initially optimistic about the Rollup-centric roadmap, as I believe any rational person would recognize price discrimination economics and that it will operate in parallel with L1 expansion.

Users with high value and price insensitivity will use L1 because it offers reliability, security finality, and traceability. L2, on the other hand, focuses on edge low-income users whose transactions are priced out by L1 fees. Therefore, Ethereum will still receive substantial sequencer rental fees.

However, the leadership of the ETH network has repeatedly emphasized that L1 as the Application Layer has actually become extinct and can never be scaled. Therefore, users and developers have made very rational responses, and the L1 application ecosystem is now dying, and the revenue of the ETH network is also disappearing.

If you think the long-term valuation proposition of ETH is a form of currency asset, then this may still be possible. Let more people own ETH and make it a form of currency. Subsidizing L2s and accumulating the value of the underlying layer to zero should help achieve this.

But if you think the long-term valuation of ETH is the widespread use of network equity in the protocol (I think this is more likely than ETH as a currency), then you need to accumulate value.

It is obvious that we really messed up this due to incorrect economic assumptions.