In D.C as frequency is zero the inductive reactance offered ( Xl=2pai(f)L)is zero and Cepecitive reactance offered is high (Xc=1/2pai(f)C) hence would block D.C. In inductor the the rate of change of flux is praposnal to change of current(L= do/di)hence inductanc could filtered the current.....
If the question is applied to a non-isolated buck/boost regulator, I would say this has to do with duality [see "Duality (electrical circuits) on Wikipedia] where voltage and current swap roles, inductors become capacitors and vice versa, etc.
In general, regulators are voltage regulators; i.e., voltage input and voltage output. If the circuit dual of the buck/boost regulator can be created, then a capacitor could be used for buck/boost of a current input and a current output.
For a circuit to have a dual, I understand it has to be able to be drawn on a plane with no "crossovers". (I've wondered whether a 'loophole' could be to draw a circuit on a topological surface with one or more holes, like a toroid with extra 'handles', such that crossovers could be avoided, but haven't explored the idea.)
There were a couple of papers on a 'current doubling rectifier' at the first High Frequency Power Conversion conference in Virginia Beach in 1984 (one was on a patent), but Rudy Severns pointed out that is was just the dual of the full wave voltage doubling rectifier. Intrigued by this insight, I explored the current multiplying duals of other voltage multiplying rectifiers in the two attached papers.
For an isolated buck/boost converter, however, we have a problem in that there is no dual for the isolation transformer. As I (informally) discussed in a few magazine articles back in the '80s, the equivalent of mutual coupling through electric fields, instead of magnetic fields, would require "currents" of magnetic monopoles, which haven't been found yet in our physical world. (See columns 9, 11 and 13 in the attached collection)