Core routines improvements

[nicamo]

• Simplified and improved the way psi_boundary is identified for limiter plasmas.
• Added new algorithm to identify the plasma domain that does not rely on the default routines in critical. This is more expensive and currently only gets triggered when the default routines fail
• Removed reliance on the Givens rotations in the NK solver, as those did not allow for regularization. Introduced regularization specifically for collinear terms.
• The regularization coefficients can now be fixed for each solver (static and dynamic) independently. This requires a small change to the API of the "stepping" object, which now needs to be provided with a static solver object (rather than instantiating its own like before)
• Fixed the change-of-basis procedure in the mode decomposition procedure

[kpentland]

This PR looks good - I've made a few minor updates myself:

We have added a new example notebook (number 10) on how to compute vertical instability timescales/growth rates and the associated quantities of an equilibrium. In particular, we:

• explain how FreeGSNKE calculates the deformable growth rate for a given plasma equilibrium.
• demonstrate different ways to visualise and plot (un)stable modes, Jacobian matrices, and how the plasma deforms under an unstable mode.
• explain how to calculate the deformable) inductive stability margin parameter (from Portone).
• explain how to calculate rigid (non-deformable) metrics such as the Leuer parameter (e.g. ratio of stabilisng to to destabilising metal forces).
• explain how to use different options to decompose (i.e. omit high-freq) passive structure modes (saving runtime).

We have also stabilised these growth rate calculations by:

• squashing a bug that was returning multiple unstable modes under some parameter combinations.
• removed some numy.linalg.inv calculations (with numpy.linalg.solve) that were likely leading to numerical issues/instabilities.

We have also re-numbered the examples to account for the fact there are now over ten!

[kpentland]

I'm happy with the changes we've made here - ready to merge.