save as pdf vs save as postscript

Postscript is for very specific laser printers.
PDF works with any printer that supports the truetype fonts that are incorporated in PDF. The difference between the two is best described here:

http://en.wikipedia.org/wiki/PostScript
http://en.wikipedia.org/wiki/PDF

Since postscript itself is inherantly better raster quality than PDF, I would suspect postscript converted PDF to be better quality. I would try an experiment and see if the file from a postcript converted to pdf is the same size as a direct PDF file. If one is larger, one logically would have more information than the other.

That's a good point about lossless and lossy compression.
The real question becomes are you trying to convert vector into raster, or not. If you are, you do lose some information in the process, because vectors are always approximated by raster. While raster regions can have inside and outside topology that is converted into vector polygon of different topology inside and outside, you don't have direction. Vector back to raster you lose the direction of drawing of vectors if they go from point A to point B.

If you did a pure postscript print job, you'll likely gain more information than PDF to postscript or vice versa. A software called Ghostscript I've heard has a much better rendering of postscript prepress than just PDF. Though I've never tried it myself.

http://pages.cs.wisc.edu/~ghost/

I use Acrobat Distiller for most of my PDF creating so double-check what I am about to say for accuracy. It may not be 100% correct. 🙂

When you save a PDF OS X is applying some compression and downsampling of any images in the document. When you print to postscript, all image data is being sent as is and when it is being converted to PDF, I believe it retains all the image data.

You can use the ColorSync Utility to creat a custom quartz filter for a higher quality PDF. At least I think you can, I never tried this myself.

-kpluck