A fairly recent example of the first: I had an old Judges Guild module "Druids of Doom" kicking around, and for some reason pulled it out and gave it a longer look. On reading it through, I thought "hey, I can use this - it fits into the campaign without me having to do much, it doesn't need much conversion, and it's got enough different elements to keep things interesting."
And so I ran it last year, and it ended up being far more DM-side work than I expected mostly due to little things I missed when I read it through: the maps and room write-ups often wildly disagreed as to room and hallway dimensions (to the point where I really think the writer was using one scale and the map-maker another, offset by about 1/3), the dungeon occupants (mostly humans) didn't make sense or tie together nearly as well in play as it seemed they'd do on reading, and some of the specific elements that read as being cool and interesting just didn't work out in play.
* - and I'm not talking abut edition conversion here (that's a separate thing), just what it takes to make a module a) fit into your campaign and b) runnable at the table.