Why does fuel flow go UP when leaning on the ground?

I observed a bizarre thing this morning. We were in a 200 HP Arrow-II
equipped with a JPI FS-450 fuel computer and were talking about leaning on
the ground to prevent plug fouling.

With the engine running at about 1000 RPM shortly after startup and the
brakes on, we were drawing 2.6 GPH. EGT was about 1000 F. As we leaned,
we got to the point where the engine started to run a little rough, the EGT
went up to about 1100 F, and (here's the strange part), the fuel flow rate
went UP to 3.2 GPH. Does this make any sense? Why would leaning make the
fuel flow rate go up?

We ran through a couple of rich-lean cycles on the mixture, and the
readings were repeatable. What's going on here?