The differential pressure at normal wind velocities would be very, very low.
If I were you, I would opt for the "spinning cup" version. When you look at all the little monitoring stations the government has put around in the national parks, all have spinning cups. And as I mentioned before, I have built such a device that worked - for awhile. My only issue was the fragility of the cups. I like Dave's idea of using old telephone bells. They are a bit heavy, but - being brass,
they should be easy to drill and solder.

And again, I used a 5.25 floppy disk drive motor. It had two good bearings, and most importantly, it had a tachometer output. I decided to use that as my starting point, since I used to work extensively with those disk drives (shows my age). You should be able to find some of those drives in a junk shop.

A few brands (Siemens for one) had a pure, amplified Hall-Effect tachometer output. You could the turn the motor shaft very slowly, and it would still put out a 5V p-p signal. Most of the others just used a separate winding.
You have to use a comparator to get a 5V p-p swing from those. And most of those motors will produce a tachometer output WITHOUT powering the motor. Just try a few different brands. The brand I used was Tandem.

My setup got thrown away when I moved from Iowa to California 19 years ago, otherwise I'd send it to you.





At any rate, the motor had great bearings