Allow for treasury.voidSpend on the respective track where the payout is approved

[DrW3RK]

Dear Fellowship,

As it stands now, the "Treasurer track only" way for voiding payouts is not working well, as it requires around 43 days even to start the process (more context: https://polkadot.subsquare.io/referenda/1800#3)

It would be great if we could cancel the payouts through the same origin that approved them. This ensures approval and voiding payouts are subject to the same approval/support curves.

[gabrieljaegerde]

This makes sense to me with one caveat.

With the Treasury Guardian we are using the voidSpend in conjunction with new treasury.spend calls to also be able to "delay" spends. What it really does under the hood is bundle a voidSpend and a new spend creation (with a later validFrom block than the original scheduled payout) all into one proposal.

I am not suggesting we design OpenGov around the Treasury Guardian tool ofc but if we wanted to retain the ability to "delay" then with the current runtime there would be an issue when USDT/C <-> DOT rates get adjusted. If the ratio gets adjusted down from the current 1DOT:10USDT/C for example then the original track may no longer have the right permissions to treasury.spend the USDT/C amounts being "delayed".

Given what the treasury guardian is currently doing to delay spends is a bit "hacky", a call to delay a payout might be worth considering if we are interested in maintaining the ability to delay a scheduled payout and want to implement what @DrW3RK is suggesting above or have a separate track just for payout cancelations/delays.

[Tomen]

related: polkadot-fellows/RFCs#159