CURA-12164 Fix innerwall seam position

[wawanbreton]

This PR actually contains a few different works:

• Rework of the score calculation for the seam point selection. Previously, the calculated scores were inconsistent and depending on each other when they should not, making it very difficult to add new scoring criteria. The scoring is now more consistent, hopefully easier to read, and much easier to tweak.
• The overhang avoidance for the seam placement was previously done after the actual seam placement, making it impossible to calculated the seam for inner walls related to the correct outer wall seam position. The overhang avoidance is now integrated in the global score calculation for the seam placement.
• The seam calculation strategy for inner walls is now forcibly set to "shortest" when outer walls are printed first, so that the starting point of the inner wall will be located as close as possible to the starting point of the outer wall.
• Rework of the user-set position for the seam, which is now treated as other strategies so that we can find a proper fallback point when the original one is on overhang.

:warning: This introduces some behavior changes of the existing seam selection strategy. Although they are more logical (no more depending on hidden settings) and more explainable, they do change the results of situations people probably got used to.

[wawanbreton]

Possible performance regression was detected for benchmark 'C++ Benchmark'.

I do not see such a difference on my PC, and the code being run for this test hasn't been touched. For now I'll assume this was a hiccup of the runner, and will keep an eye on this.