Also changes the [Output]

* Move coefficient computation into RuinaNonlinearity
* Allow some friction data to vary in space

x/|x| isn't that big a deal and the logic was a gigantic mess

As a consequence, the nonlinearity is only assembled once

Also kill the kink logic and thus minimisation2

Kill tests that use the old interface

Drop members a, V_0, K
Move a into coefficientProduct
Move V_0 into V_m

Add comments that show what we're computing

Do not re-create overallSolver for every timestep