I am getting close to 1000 characters and can feel the power of the character components working on my brain. "Chunking" is a well-known memory technique where you encode several facts at a time using other previously-known facts; in the case of Chinese characters the chunks are the reused components, whether they be radicals or characters in their own right. When I already know the components, learning a new character is much, much, much, much easier, but often I don't know the component and so am stuck learning a new component together with a new compound character.
That suggests to me that, in order to efficiently chunk characters, you would need to study them in order starting with the basic components and building up to the compounds that use those components (à la Heisig). This is not currently how Skritter works, but it could be done manually with custom lists.
I'm wondering how you plan to design the component study system, and if it would be possible to have a "study components first" option for the scheduler that would notice that character 辆 is made up of 车 and 两 and then queue the components for study before we study 辆. I am referring here to initial exposure to the character, not to the spaced repetition review schedule.
James