All my filters are canisters, so they transplant easily. Condition a newly set tank with a single newly set canister fishlessly. Move canister to another freshly set tank. Challenge both the newest tank with the conditioned canister and the "cycled" tank with with an uncycled filter added. Test ammonia/nitrite/nitrate after 12 and 24 hours. Guess where the ammonia is always oxidized to nitrate within 24 hours? The tank with the conditioned filter, not the conditioned tank (which almost always showed less than 10% of the ammonia challenge gone within 24 hr). I have difficulty testing to 10% repeatability without using the LaMotte tests, so I have to dilute down to the range suited to those tests.
Repeat until you sure it is not fluke. It is not.
Corollary trials from planted tanks: 1)Tanks which are heavily planted and also equipped with biofilters do not rely on the biofilters as much as the plants. If you "borrow" a long-term filter from a healthy planted tank, put it on a new tank and ammonia-challenge it, it will not be competent to handle the bioload of the original tank. But, the long-term filter will mature when fishless is continued in a very short time, as it is well inoculated with the needed bacteria. 2) If you can borrow a comparable filter from a FO tank, use that on a new setup, ammonia challenge that setup and it will be fully competent. Conclusion: Plants out-compete biofilters for ammonia when they are healthy, but they do not necessarily get all the nitrogenous wastes, the bacterial filter gets some.
Try it for yourself. Replication of experimental data is valuable.
Another corollary: It helps to run a lot of similar tanks and a lot of similar filters - it makes "borrowing" much easier.