e.g., if a page is http://web.archive.org/web/19980530120000/abc.html and I use URL substitution to replace /199805 with /199805/ (making a deeper directory level), then when I go to download at a later date, OE tries to download http://web.archive.org/web/199805/30120000/abc.html
This kind of URL substitution is necessary because the "Prevent download directories from overloading" feature only applies to too many files in a directory, not to too many subdirectories in a directory. Since Internet archive puts the full timestamp as a directory name, this would result in millions of directories being created inside of one parent directory. That seriously overloads Windows.
I am sorry, but I don`t have a solution to that URL Substitutes update problem. I could record original URL of a saved file somewhere, but it may happen that URL Substitutes rules will be used to make two simular files from different URLs. In this case, it will be impossible to understand, which exact URL was used to load the file.