I guess option 1 isn't worth investing in, then. But compression (option 2) could still help considerably, even for files of a few megabytes. I imagine it wou

I guess option 1 isn't worth investing in, then. But compression (option 2) could still help considerably, even for files of a few megabytes. I imagine it wouldn't be easy to change what's in the files, but a 3rd option would be to look at what they may have in common. For instance, if the user needs 1000 files that all have a very similar first page, there may be some way to avoid making 1000 copies of that information.