Had access to a commercial 3D printer at a hackerspace. After the initial few cool test objects, the annoyance of dealing with broken STL files and waiting hours for tiny fragile parts overtook the "convenience". But I think those will both improve with time.
My sub-poll question would be: for those who do have a 3D printer, do you create your own content or print only third-party STLs?
In my opinion, current generation 3d printers have to be looked at as tools, not a holy saviour of manufacturing industry. It has it's strong points (fast turnaround times, low price) and has it's weaknesses (subpar resolution, support material is required for many types of parts, being finicky).
Nonetheless, I'm really happy with mine- I built a RepRap around 2 years ago. It took full 1 year to get from starting to read about it, to first prints good enough to be useful. After building, it gathered dust for some time because I didn't know what to do with it- the process of building was more interesting than process of using it :>
That being said, I'm also experimenting with DIY SLA printers- by using UV laser and rotating mirror assembly (found in 2d laser printers). Don't know if it works yet- I'm finishing up electronics/firmware and trying to come up with mechanical design.
I have a small printer on the cheap end of the spectrum. I started off printing things from Thingiverse because I was having a hell of a time trying to create STLs from the 3D modeling software I already knew (Lightwave and Maya). It worked but my toolchain was a huge pain and the resulting geometry was not always sound for slicing to gcode.
In recent months I started using openSCAD and now almost exclusively print things I generate with it. I still check Thingiverse to see if people have already created things I'm attempting to make, which I see as the real power of this whole experiment.
My sub-poll question would be: for those who do have a 3D printer, do you create your own content or print only third-party STLs?