How about you make specific predictions (written) of what will happen if you abstain for a specific number of months, then abstain for the given number of months, and then evaluate the original predictions?

For things like "clarity of mind", find some way of measuring it. For things like "motivation" instead focus on what exactly you will be motivated to do.

Then compare with the same amount of time with porn.

It's still very little data, but better than no data at all.

Less meta -- I think it pretty much depends on what you replace it with. Which can be both someting better or something worse, and you probably don't know the exact answer unless you try.