But what if my cistern is full already?

Non-structural BMPs (raingardens, infiltration basins, etc.) should be modeled as dry, and able to dewater within 72-96 hours. Consult with your local jurisdiction on their dewatering requirements. Cisterns are a tad bit different. You will need to determine the demand for water reuse and size appropriately. If you are using for reuse, your cistern will likely be designed to hold a much greater volume of water than that which is generated by the low-level storm events. Your cistern should always be equipped with some form of bypass as well.

Thanks Michael for the reply. I understand the functional need for a cistern larger than one used strictly to capture the single rainfall event, but I am very space-restricted because of the cistern location and am wanting to be certain that what I install for storage capacity will meet requirements for the credit (I'll only have one shot at it; adding on if I undersize won't be an option). I intend to reuse for toilets and cooling tower. My 2-year, 24-hour rainfall capture requirement is 3,000 gallons. Can I assume a dry tank at the start of the rain event to qualify for the credit and install just a 3,000 gallon tank? Or will they require larger capacity, based on the assumption that there is already at least some amount of rainwater in the cistern before the rain event begins? Lastly, if they DO require more capacity, how would I know how much more? Thanks a lot!

Ethan are you referring to V4 or 2009? I am assuming 2009, because you reference the 2-year 24-hour storm. Calculate the delta volume for pre and post 2-year storms. That is what you need to meet the volume requirement. As for rate you need to develop the hydrographs and show that reduction. Keep in mind, I'm not a GBCI reviewer, and this guidance is only that. If you need this to certainly comply, I suggest you contact GBCI via their inquiry page and discuss this with them.