Thanks for the suggestions! I wasn’t aware of the internal pull-up resistor. I reconfigured it to simply wire the meter between GND and D3, but this ended up leading to lots of false pulses. The meter is about 80’ away from the emonPi, so I must be picking up some noise in the wire. Adding a 10K resistor between 3.3V and D3 seemed to fix the issue by giving it a stronger pull-up.
It appears this setup gives me clean pulses from the water meter and no longer interferes with the energy readings. I’ll keep an eye on it for a few days and give an update if I run into any more issues.