To mappack or not to mappack?
Up to you really.
If you want to make a small campaign and are up for it you should. But if you cannot dedicate yourself to a project and like to do it in small maps do it that way.
However creating a mappack shows dedication and skill = respect.
If you want to include some kind of intertwined continuity, then I suppose that's another story. Still, I know some people have opted for map "series", rather than releasing all at once.
I'm working on one now, but as others have said, it does take some considerable time and dedication to do 
Rob.
Now it's still double the amount of what you think it is.
It is quite time consuming, but I believe the pay-off is worth it.
What I do however, is letting some friends play through the levels to gather their feedback, improving maps, and posting some of the maps under work in progress.
I think that a mappack can submerge people into your maps, puzzles and willing to solve 'em all! Where a single maps is just done after an x-amount of minutes 
~Lp
I also think that a lot of mappers just started with a few single maps, and later starting with mods/mappacks. And the developments of mappacks/mods is just a lot lot longer.
On a side note, how long does it take you guys to finish a map? Its taken me about three weeks to finish the one im working on, and about half of that was just detailing, am I spending too long on one map?
At that point I'll upload my maps to the WIP section and wait for some other feedback (also takes quite a while.) Make some other final adjustments and I'm done!
So I guess the whole process takes me 2-3 weeks. A few things I do to make mapping faster:
* Use two screens, one with a windowed mode portal 2 (no border) so I can easily alt+tab. The other with hammer
* I don't add lighting until the puzzles are working.
* I don't add nice texturing until the point where the map itself is good enough.
* I don't add elevators until the puzzles are working
* I run maps fast/fast so that compile times are lower
All those things make it easier to move objects around and/or make compiling times lower, which alows you to playtest more in the same time
EDIT: good luck with your mappack 
I agree with almost everything IPfreaky said. Though...
Planning is a very important stage in the development of a map. If you haven't started mapping and it has been two days jotting ideas and making rough sketches, I think you would have more intention on completing the project than someone who starts Hammer immediately.
Lighting is important as well. What I do to save time is use Cordon to compile certain spots of the map to test lighting, instead of waiting 5 minutes each time I want to test one light. (Try radius culling too)
I didn't know about cordon, but I will use it for my next map! thanks
My first map is coming out soon, since I've never released a map I think i'll ask for some feedback on that before putting much effort in a bigger project though.
BrainstoneX wrote:
It's at least double the amount of work that you think it would be. right. now. at this moment.Now it's still double the amount of what you think it is.
Hofstadter's law?