I didn't find this problem for latest version of Multiwfn (the one updated on 2020-Jul-11). Please check the release date when Multiwfn boots up to make sure that you are not using old version.

Best regards,

Tian

From last time, I wholeheartedly appreciate that you have addressed the underestimation of bond path length. It seems your latest Multiwfn version suffers this problem again.

Here are my calculated results using the latest version of Multiwfn:

c30 c55
The distance is    4.276803 Bohr
#  121     CP:  172 (3,-1) ---> CP:   30 (3,-3)   Length:  2.73000
#  122     CP:  172 (3,-1) ---> CP:   55 (3,-3)   Length:  1.50000

c55 c54
The distance is    4.238525 Bohr (    2.242931 Angstrom)
#  133     CP:  180 (3,-1) ---> CP:   54 (3,-3)   Length:  2.67000
#  134     CP:  180 (3,-1) ---> CP:   55 (3,-3)   Length:  2.46000

c30 c54
The distance is    4.544420 Bohr (    2.404804 Angstrom)
#  109     CP:  164 (3,-1) ---> CP:   30 (3,-3)   Length:  2.22000
#  110     CP:  164 (3,-1) ---> CP:   54 (3,-3)   Length:  2.25000

I await your response.

Sincerely,
Pregabalin