M100 had numerious changes and quit working.   Part of the problem is
the overloading of the SERIAL_PROTOCOL functions.   Also, some of the
address arithmatic was changed to use char *ptr and passing ptr into the
SERIAL_PROTOCOL functions caused them to try to print a string instead
of a number.     M100 is working again.   Let's keep it that way!

M100 has been expanded to now have a function  int
free_memory_is_corrupted()  that can be called from other code to see if
the free space is still contiguous.  It may make sense to add a flag to
control its verbose nature but right now, the extra chit chat is very
helpful to know int free_memory_is_corrupted()  is doing the right thing
and what it found at various points when it was called.     A 'Show &
Tell' is coming up with int free_memory_is_corrupted().

When PARK_HEAD_ON_PAUSE is defined Marlin won't compile because of two missing semicolon.

These fixes are neede because before it showed the englisch text even if language = de was selected

Firmware was forgetting the z_fade_height in Planner, set by M420 Z<f>,
after each reset or loading settings from EEPROM. Added the float
z_fade_height to EEPROM-Storage, now remembers the value.

The `ftostr4sign` function got munged by the `WITHIN` patch

Add support for new electronics
www.reprapobchod.cz

Having these items randomly (and errorniously) disappear cost me several
hours of work yesterday.

DAC_STEPPER_DFLT wasn't changed to DAC_MOTOR_CURRENT_DEFAULT in all
places on PR #6124