" wrote:

An amp meter measures amps, not watts. Even if you assume 120V,
you're still calculating volt-amps, not watts.

A volt-amp is a watt, when the load is resistive.

If I perform high-speed sampling of both the voltage and current, and if
I multiply each reading together to get VA for each sample, and if
integrate those VA samples over time, I will get the actual watts or KWh
that I should be billed for. That would correctly take into account
reative / inductive loads (like motors, light ballasts, computer power
supplies, etc).

If I simply calculate watts as equal to VA based on the current
measurement from a clamp-on meter, then I'm over-estimating what the
billing meter is "seeing" because I'd be assuming that all my loads are
resistive. In other words, my calculation of watts = VA can't help but
assume that current and voltage are in phase with each other.

The billing meter knows how to calculate wattage correctly when the
current and voltage is out of phase.

I guess a clamp-on amp meter that also had a couple of voltage probes so
that it could simultaneously measure the voltage could measure true
wattage would be needed.

The errors you mention above will likely be even bigger, though
Surprisingly, a watt-hour meter (it's already there) is the real
way to measure watt-hours. ;-)

As long as the watt-hour meter is working correctly.

Which seems highly suspect given all the info I've been posting here.