Forgive me for sounding like a broken record, but “crypto values” like “decentralization” are merely a means to an end. All hypotheses and conjectures don’t matter. The only thing that does is - are these feature improving consumer experience and choice?

Let’s take crypto’s #1 usecase, as an example - alternate store-of-value. Decentralization is certainly a key ingredient to this usecase. At the same time, a vast majority of consumer opt for custodial services. This is perfectly fine, we should all embrace choice, instead of gatekeeping. The right way is the best way for each person, and for most people it’s not onchain.

Conversely, crypto’s #2 usecase, stablecoins, are defacto centralized. Yes, it’s a spectrum, but we have nearly a decade of evidence now that there’s a rough inverse correlation between demand and decentralization. USDT and USDC command a majority of the market, with semi-decentralized solutions like DAI being a much smaller player, and maximally decentralized solutions like RAI or LUSD struggling to find product-market fit.

If decentralization does not actually improve the consumer experience, it’s not that useful. Sure, there will always be a tiny niche that will use a RAI or LUSD, and that’s perfectly fine for consumer choice - however, it’s important to prudently focus on solutions that do have product-market fit.

At the same time, it’s very important to understand that the best parts of stablecoins aren’t actually about decentralization - very evident from USDT on Tron still being the #1 consumer onchain application, and most of said users using CEXs as intermediaries to interact with Tron is another. Still, I do believe USDT/USDC etc. on Ethereum L2s will be a marginally better consumer experience, because a) they are more efficient, reducing cost to the consumer; b) increasing resilience against attacks and failures; and c) much higher probability of long-term sustainability.

This brings me back to hybrid applications, once again. To offer the best consumer experience, and efficient operational costs - that are passed on to the consumer - you don’t need to “blockchain everything”. Indeed, you should only use blockchains for the elements that do actually improve the consumer experience in some way. It may be more costly and less efficient, but it may be a critical feature that makes the app special, so it’s well worth it. It’s just one potential feature out of many. For everything else, it’s perfectly fine to be “centralized” - if it leads to better consumer experience and choice.

Which is not to say that some applications don’t benefit from being 100% onchain - but it must be justified with adequate product-market fit given the operational and development costs.

In the end, that’s all that matters, good, sustainable applications that add value to the world. Decentralization may or may not be a useful feature - in most cases it is not - and application developers must make prudent design choices accordingly.