[Fish]

I'll try to dig up the link where I read it, but someone actually did a study and caught all their return water in a micron sock and then examined all the pods. Their findings were that the pod-vs-impellor mortality rate was very very low. It was something like only 5% of pods are killed by going through the impellor.
I don't know where the idea that a refugium should be low-flow came from, but I personally don't see the benefit. My 20gal display had 40x turnover and it was crawling with pods. If you have pods that are being blown away, then you might have a velocity problem, not a turnover problem. Infact, have you ever tried to purposefully blast a pod into the water column? You have to sneak up on it with a powerhead to do it.
You may have to change the name to "sump-with-liverock-chaeto-and-pods" instead of a "refugium" to make people more comfortable, but as far as I know, it will be healthier with a higher turnover than it will with a lower one - just like your display. In fact, my 20gal display had a 40x turnover because of the sump return and a closed loop - but my sump only had the return pump driving it so it's turnover rate was half that... so what did I do? I added a small powerhead to the sump. The LR and chaeto didn't "know" it was in a refugium - so it still benefited from better water movement just like it would have in the main tank. ( :
Haha I guess an extreme example is my cannister filter mod that turned a 1.6gal eheim 2028 with an output of 250gph into a "refuguim". I added lights, LR, and cheato to the inside of the filter and it became a breeding ground for pods - despite having a volume turnover rate of approx 156x!!!
- Chad