The parameter order was wrong on mark(), marked() and unmark().    

This was breaking the G26 Mesh Validation algorithm.

Co-Authored-By: InsanityAutomation <insanityautomation@users.noreply.github.com>

tried to fix it with snprintf in libs/duration_t.h but its worst... :/

The LCD Display shows 3 digits of precision, not 4.   
So change the bounding value to something more rational.

Allow LCD Menu to adjust Junction Deviation lower than .01.   Some machines currently have a JUNCTION_DEVIATION_MM value as low as .005 mm.     In the case of non-Linear-Advance machines...   The upper limit is raised to .5 mm.

Probably more thought needs to be given to the Linear Advance case.   It maybe it doesn't need a special case and should just use the same bounds as the non-Linear Advance case.

Reassign FIL_RUNOUT_PIN from P1_18 to P1_19

P1_18 is connected to a buffer in order to output 5V making it unsuitable for input.

The extra clutter was making the Junction Deviation patch script calculate the wrong value.
(The script wasn't smart enough to see the extra lines were commented out and used the wrong values in the calculation.)

All fixed now with a JUNCTION_DEVIATION_MM calculated at .017

Re-apply the formula .4 * V^2 / Default_Acceleration to work correctly in Raptor configuration.h file.
(The problem is there are multiple XJerk and YJerk values declared based on the step sticks used on the printer.)
This change should provide correct values for all setups.

Re-calculate and fix the Junction Deviation value using correct DEFAULT_ACCELERATION number.