That's kind of my point. There is no complete or prefect methodology, per say. "Refugium" is an inadequate catch all term for a wide range of somewhat experimental applications. There are few hard facts, a lot of disagreement, and the book has yet to be written.
Unfortunately, I've noticed a reduced interest in refugia lately. In particular, the slow-flow design is now often overlooked for the simplicity of inline tumbling chaeto. But I find this far too limiting in application and question the overall usefulness of chaeto when compared to DSB and GFO alternatives. I prefer pod production for a supplemental food source and believe when combined with a slow flow, it improves biological filtration of detritus. I use a DSB in the display for nitrate reduction. Some use a DSB in the fuge for that, but I believe it to be much less effective and more prone to problems.