I, too, had to stop listening after about thirty seconds, because I couldn't understand a word (I was in my car).
HOWEVER, I continue to be in favor of Pseudopod experimenting with production. I've been doing quite a bit of production work lately, finishing up an audiobook, and I have a few suggestions:
1.) I think the reason why the other stories mentioned didn't draw many complaints is that the sound effects used didn't actually interfere with the narration itself. Noises with a constant, undulating presence (like blowing wind, falling rain, etc.) don't tend to take focus away from the vocals, because the listener's mind can tune them out to some degree. Basically, the mind thinks, "Oh, I'm familiar with this noise, and I know it's going to go on like this in the background for a while, so I won't pay too much attention to it." In that sense, they enhance the listening experience without getting in the way. Noises with frequent discrete events (like voices speaking recognizable words, footsteps, or environmental sounds that start and stop) tend to be distracting, because the mind has to take focus off the narration to recognize and evaluate the new sound. This will often cause the listener to lose track of the story.
2.) When the vocals themselves are processed you have to be very careful. One trick I've learned to deal with this issue is to create two identical vocal tracks, pan one to the left and one to the right, and then put the effects on only one of them. Whichever one gets the effects, lower the volume on that one slightly so that the un-processed one is dominant. That way you still have a clear speaking voice, but the listener's mind recognizes that something additional is going on (robot, demon, cybernetic rabbit, etc.).
Hope this helps!
P.S. Thanks for the Candyland references. That made my day.
P.P.S. Has everyone seen The Cabin in the Woods yet? Go see it. It's phenomenal.