I thought the variable capacitance input was a cool idea, but decided to change it firstly since constant inputs have not yet been implemented. This means that if you're using the component as part of a flip-flop design (as a random example) then you'd either need every flip-flop to have an extra input to control the pulse width, or you'd have to 'hard-code' that input out of nand gates, both of which feel a bit clunky.
I could have waited until I've figured out how I want to handle constant inputs, but then I also didn't see a huge advantage to having the value be logic-configurable. My thinking was that you want to have the pulse width set to the minimum value that gives consistent behaviour with your design, and so once you've calculated that, it would be something that's typically left constant rather than configured during operation.
I have no doubt that some use-cases exist where the variable input would be more helpful, but I think that the way it is now is most convenient for the most common cases.