I agree, if we need an upgrade, build a proof of concept wallet/tool on liquid with all the tools available there now. Then when it is clear how valuable it is we can port to bitcoin.
Discussion
Did you open the link shared above?
Yes. There are many cool uses there. Like I said, a use is not a solution to a problem. I would just like the problem to be stated. Is the problem with bitcoin that there aren't payment pools and that catastrophic failure going to ruin the technology? Or is it a feature(use) that people would really like? This is my concern. That these things are simply being added because they would be nice NOT that they are integral to bitcoin's success.
Everyone who uses bitcoin and pre-signed transactions, understands the problem these use cases are trying to solve.
They improve privacy, security and UX and the list is not complete. You are free to avoid using them after soft fork activation. Nobody is forced to use taproot either.
So to parse the answer out of the implications that I don't know about the privacy concerns: if privacy is not achieved on the base layer the asset fails as a monetary asset and ultimately is not useful as a global value store.
I personally see privacy as a function of medium of exchange (where it matters most) which is firmly in the layer 2 application. We've had two cracks at enabling layer 2 compatibility. Both have had knock-on effects and unintended consequences. I believe CTV is no different. But the trade off seems so short sighted, I can't believe no one is looking down the road that it would pave. "Shared UTXOs" have no economic analogue beside some communist ideal. You either own something or you don't. But that's an entire tangent.
The point is you don't need to change consensus rules to maintain the underlying asset (beside the Unix time code parms) to enable more private payments, transactions, or layer 2 functions. The calls made to the base layer should only transfer UTXOs from one set of addresses to another. All other functions can be abstracted to the second layer. Look at it this way: [0,1] <-you don't need to change this to do complex calcs in a system. You just need to build hex on top of it and assembly on that, and language on top of that. Stop trying to alter the rules of binary to make things easier, it might break everything already built on binary.