When using the in-place-patch method the default minimal shift size seems to be 2 times the segment size. Why does this not default to 1 segment?

To allow using the flash memory to include the step data at the end of the flash region it could be advantageous to allow specifying some extra space that is taken into account when calculating the shift size.

As an example adding 1kB to the from_size would create a writeable area of 1kB at the end of the memory region that could be used to store the step data. Would such a change be accepted?