This was going to be an alternative implementation of BitEscrow, the extended musig signing is designed for computing adapters that attest to execution of a state machine.

However you would need a reliable group of oracles for time and for outcomes, and then you still need arbiters to handle disputes for edge cases, plus signatures from other members of the contract.

Maybe in an ideal future, there's a healthy decentralized market of oracles that you can use to represent the conditional gates you need for the machine. Maybe oracle servers can be lightweight and ubiquitous, and the process of enrolling them into a contract would be painless.

In practice though, this is a very fragile system, because you are relying on so many intermediaries as oracles in order to decentralize the attestations. Also there's currently no market for this, so you have to stand in as the oracle for pretty much everything, which makes you the custodial computer and defeats the purpose of using DLCs altogether.

Lastly, DLCs only work for computation that is deterministic, because you have to pre-compute all execution paths in the contract in order to create the adapters. Things like time and price-pegs have to be modeled in complex and convoluted ways, in order for them to be usable in a DLC based machine.

In BitEscrow, we have a far simpler system, we can support any kind of computation for virtual machines, and the tradeoffs are marginal since we still use covenants and provable execution.

We also have a working product that you can try out today. Our API is available on three separate test networks, with full documentation and live demos. Plus we have a complete SDK for building (and offering) your own smart contracts:

Our code is open-source, so if you want to ditch our contract VM and hook up your own, you can do that too. It's pretty easy and I'll help you out: