If the clip is a long one - lets say 45 minutes (half of a play or concert), HDD will break the file into 20 minute clips (2 @ 20 minutes; 1 @ 5 minutes); MiniDV tape will be one continuous file. There are no lost frames between the HDD file segments, it just "is".
Generally, the video editor can help with the segmentation. In the HDD environment, you would stick the files back together to make a single continuous clip for the project. In miniDV tape land, while it is true that importing the whole tape (even with starts/stops) results in a single large project file, you can make the short clips by exporting the short portion desired or just import what you want.
In the case of miniDV tapes, the tape is the archive. While it is accurate that importing real-time can be time consuming (1 hour of standard def takes 1 hour to import; hidef may take longer that real-time and depends on your computer's CPU), I find it easy enough just to go do something else... design the DVD art, mow the lawn, whatever... In the case of the HDD archiving, once you copy the files from the camcorder to your computer, your first step *should* be to make your archive files. When the project is complete and you delete the files from the camcorder and the computer... and 3 years later you want a clip that is long gone because you did not archive those video data files... whereas, with miniDV tape, just go back to the tape and re-import. So the perceived amount of time saved in the import/copy process is less than initially thought.
There's more - like how DV is much less compressed than the internal hard drive MPEG2 compression (and the compression can impact video quality) or dropping/breaking the camcorder poses different video recovery methods or the differences between running out of hard drive space or miniDV tape will impact your planning or reaction requirements...