I've actually been timing all my work using an app called Grindstone.
On pl_snowycoast, my hours invested breakdown like so:
Total time to reach Contest Ready (B1): 88.5 hours. - This is the time it took to go from an empty Hammer file to the (buggy & slightly broken) contest version B1.
Total time to iterate B1 -> B2: 4.3 hours. - This was an easy one. It's just the time it took for me to fix the obvious big problems, do a few correction passes, test it, and pack it up.
Total time to iterate B2 -> B3: 66.5 hours - This was quite big due to being when I created the majority of custom props for the map.
Total time to iterate B3 -> B4: 53.1 hours - The big time sinks in this one were the spaceship (14 hours) and the cinematic props (27 hours).
I also include in that the time it takes to make promotional images, build cubemaps, etc. Right up to the point of the bsp being packed up and ready to go. I did not however include time watching demos.
It takes me about an average of 12 hours from blank file to uploaded for my workshop models (weapon or hat). In that time, texturing always takes significantly longer than the modeling itself. So 14 hours is not too far off the curve for the spaceship, and I did spend a lot of time trying out different engine shapes before settling on the current design.
The cinematic timesink was mostly due to experimenting with the best format to export for Valve source. The simulations themselves were very quick (a couple of minutes maybe?). Rigging and hand animating the metal "stress bends" took longer since I couldn't simulate those with good results, but it was much faster and simpler than rigging a character.
Okay, so what's all this nonsense about then?
Well, to make a map like pl_snowycoast, including all the assets and assuming a 40 hour work week, it would take me 5 and a 1/2 weeks to create -- NOT including testing or watching demos.
Sooo... TLDR, I would not even be able to complete a single map in a month's time.
