I'd like to add that I've done some alcohol testing myself.

With a sample size of one, I can definitively state that a liberal application of alcohol to the tester (me) yielded an improvement of 67.2% in honing ability over the baseline of raw newbie.

I attribute this effect to being less careful.

Most likely any benefit achieved this way steadily diminishes with both experience and quantity of alcohol imbibed. Further testing is hampered by a lack of research dollars (plus razors) compounded by the experimenter's general reticence to add CA sutures to the testing protocol.