Implicit Time Dependent Fields not updated every other time step

[fractalsbyx]

@john-mcki If you have anything to add.

[landinjm]

Yeah, you can see this in the regression tests for the explicit and implicit Allen-Cahn equation.

Not sure where this might be coming from. Perhaps something to do with the solution swapping?

[landinjm]

Seems to have changed in #655, which supposedly fixed the handling of old solutions 🫤

[john-mcki]

Not much to add. Just wanted to point out that the initial timestep and proceeding tilmestep are solved, and then prisms solves every other iteration