I prefer marcoemme41's method, especially because I also occasionally notice that (using quite a few storage providers for evaluation purposes) some files are not always synchronized / stored for some reason. That cannot be checked manually, even when only storing several thousands of files for evaluation purposes.
Regarding pathname lengths, although I'm using some sort of (comprehensive and free) Linux layer on top of Windows (called Cygwin) which is a powerful environment to handle all kinds of tasks, I think I would do the following under Windows' Command prompt to get a first impression of long pathnames:
dir your_toplevel_encrypted_directory_name /s /b > filelist.txt
( dir your_unencrypted_directory_tree /s /b > filelist2.txt )
gives you all your pathnames in file filelist.txt, sort this file according to the length of lines (i.e. pathnames) and you will get an impression of the most problematic pathnames.
In addition, when using filename encryption, exit BoxCryptor and start it again as:
this way, BoxCryptor will show both the encrypted and unencrypted names together when browsing an encrypted directory hierarchy in a command shell or Windows Explorer.
Last but not least: BoxCryptor is in important ways encfs compatible, and (not considering several encfs IV chaining options) encfs only allows for a few name encoding techniques. Stream name encoding results in shorter encoded names than block encoding, of course.