fassage,
As loose-electron stated the opamps open loop response nothing like the closed loop response so it does not make sense to do so. When the opamp loop is closed the input will look like an AC ground at least until the gain and bandwidth deteriorates it to impedance greater than that.
Unless your circuit drives a real impedance then the output power is not what you are looking for. In your circuit where the load is a capacitor so you will want to observe the distortion in voltage at the output referred back to the input ( provided that the circuit is being driven by a matched source). If this block is in the interior of a system on chip then IIP3 and P1dB are not the real specs but rather an "equivalent" voltage at the input "as if" it was a matched system of some impedance.
If you are trying to save current and not designing the opAmp to meet the loop bandwidth, gain and drive strength required to achieve distortion levels from feedback then:
To simulate the distortion of a single stage of your filter you will want to drive it with current source and load it with the next stage as an out put.
Then you will need to map a transfer function of current at its input to the previous stage current when it drives it when the entire filter is connected and driven with a power or power equivalent voltage.
Finally measure the the output voltage with distortion and map it back to the filter input using the previous transfer function combined with the opamp being tested current to voltage transfer function.
You can obtain the transfer function initially by using ideal opamps with the filter R's and C's.
http://rfcooltools.com