How does flow rate of Ion channels depend on the length of the selectivity filter?

The length of the selectivity filter of ion channels such as the potassium channel commonly is only half of the thickness of the membrane. Some book such as MBoC says this can benefit the transfer, so I wonder how will the flow rate of the channel depend on the length of the selectivity filter.

Intuitively, the longer it is, the lower the flow rate will be. But consider the following factor this may not be true:

There is no "resistance" in the selectivity filter. The ions in the channel are basically in thermal balance with the selectivity filter wall and bounce forth and back randomly all the time. The driving force of the flow is not kinetic energy, but the ion density difference at both side just as normal diffusion process. From this point of view, the flow rate should not be depend on the length of the selectivity filter, at least not very much.

If so, then the fact that the length of the selectivity filter is only half of the thickness of the membrane maybe only a coincidence in evolution history and doesn’t bring much benefits.

Considering most ion channel have very similar selectivity filter length and synthesis of new ion channel proteins with longer selectivity filter is not possible with today’s techs, this can’t be decided by experiments. Is there some molecular dynamnics simulation about how the flow rate depend on the length of the selectivity filter?