The presence of a lockfile is no problem at all. If no one is using the database you can delete it but "why bother."
The way that MS-Access does locking is quite interesting. It takes advantage of the fact that the operating system allows you to lock byte-ranges in a file even though there is no such data in the file. I could for example lock byte number 1,234,567 in a file that contains .. nothing.
Access puts almost nothing at all in the lock file, but has a rigorous and well-defined structure for the various ranges of byte-addresses that are used in the lock protocol. (e.g. say "if byte #12345+X is locked, then it means that record #X is locked").
It's a really cool system if the network operating system that you're using is totally cool and totally working on the lock-protocol, and if every workstation really does get absolutely-correct results from all those lock-requests...