Thanks for the replies and the mention of "Marimo Moss Balls". I had never heard of Marimo moss balls before and they look pretty interesting.
However, I was originally not talking about Marimo, but instead I was talking about a DIY Wabi-Kusa with mud from a local river bank and moss/small-leafed-plants tied on the outside and/or planted in the mud ball.
After making my first Wabi-Kusa, I realized that, at least with the mud I'm using, the ball is a little fragile - moving it even a tiny bit in the water unleashes a "mud storm". So, I would be a little wary about running an airlift tube through the ball.
Furthermore, I guess that mud that is sticky enough to stay in a ball shape is not porous enough to let water flow through like a sponge filter. I guess that if I tried to make the Wabi Kusa ball into a sponge filter, the uplift tube - which should be gurgling sponge-filtered clean water out of the top - would instead be sputtering muddy water out of the top and slowly disintegrating the ball. In other words, a sponge filter works because the filter material does NOT get sucked into the uplift tube, but instead serves as a static barrier between dirty water and clean water. Mud, by itself, can't serve as a static water barrier and would get sucked into the uplift tube.
However, moss might be able to serve as a filter medium... hmmm....
I guess the basic idea, that I still think has merit, is that a sponge filter causes water flow through an airlift tube and thus traps detritus from the flowing water into a (non-living) sponge. This detritus is beneficial for plants. So, can't you do away with the sponge, and use plants themselves (e.g. moss, or plant roots) to trap the detritus as it flows through the airlift tube? This way you would have an all-natural mechanical and biological filter that never needs to be cleaned or replaced.
Is this possible?