The ws-320, according to the spec sheet on the Olympus web site, samples at
44.1 khz but has a bitrate of 64 kbps. This suggests a compression of 20
to 1, approximately. This will be adequate for voice recording. I say
adequate because I have a background in audio engineering and am rather
picky about such things. A bit rate of 128 kbps would be better. 64 will
kind of mess up the sibilents, the hissing sounds with letters such as s
and some c pronunciations. These are very important to
intelligibility. The 320 should produce decent quality speech, but beware
of trying to record music or nature sounds with it, you may be disappointed.
The spec sheet did not say the sample size, but doing a little math, if it is a stereo recorder, the sample size may be only 8 bits, and this generates a fair deal of noise in the signal. Still, with care and a quiet background, one should be able to make a decent podcast with this machine. I haven't looked up the ds-2's specifications, but if it has a higher sample rate or a greater sample size, also called bit depth, it would produce better sound.
The ws-320 does have a gig of onboard memory, that is nice and even at the highest quality, it will record for over 35 hours. While it would not do for the greatest quality needs, it could be a lot of fun to work with.