Time iteration consistency + restart file documentation

[j-signorelli]

• In the SU2 source code and corresponding Python examples in TestCases, the TimeIter and time variables are updated after Output is called. This PR updates the run scripts to match that, as it was for the preCICE v2 version: Revamped Adapter to Work for SU2 v7.5.0-v7.5.1, Both CHT + FSI #32.

• Dual-timestepping (an implicit time scheme) should have any the time-varying source terms / BCs / etc. computed at the t+dt state. When importing a restart file as an initial condition, TimeIter is initialized to 1, and time to dt. In general across SU2, all time-varying "things" are computed with the time TimeIter*dt. So, when importing a restart file, computations are performed with the time-varying "things" at the correct t+dt time. Additionally, output files are then representative of the solution AT its suffix iteration number (output_000001 is the state of the solution at time dt). However, if restart files are not used as an initial condition, then TimeIter is initialized to 0, and time to 0. The computations for time-varying "things" are off by a timestep, and the output represents the solution AFTER its suffix iteration number (output_000000 is the state of the solution at time dt). More information is provided at su2code/su2#2353. The punchline is that is it better to import a restart file as an initial condition always. This is also particularly important for restarted simulations. This PR adds documentation and guidance on this note.

Closes #42

[MakisH]

The changes make sense to me, thanks!

Sorry for the delay, I somehow missed the "marked ready to review".