> ​I would like to know about typical use cases.  How is this better than > just invoking actions directly from another dired buffer?  Does it make > what one might want to do when multiple buffers are involved more > convenient?  How?

Hi there,

You can see some of our discussion on the use case on my MELPA PR:
https://github.com/melpa/melpa/pull/9207

But, to summarize, the general idea is that you capture files into the list, in which you can perform actions with them LATER. This also forgoes the need
of having multiple visible dired buffers present in the first place.

Furthermore, this allows for more explicit behavior of these actions in the event of having more than two dired buffers present on the frame
at once. This also means that you do not have to rely on the implicit behavior or dired-dwim-target (which requires cycling actions, if the
implied directory is not the intended one).

Lastly, this allows for the user to perform these actions, multiple times. (In the case of the symlink, symlink-relative, and copy functions)
As in, they can copy the captured files into one buffer, then move to another buffer and copy the same set again. So on, an so forth.