Inductor Q measurement

Believe the data sheet. The real inductor is tested under real conditions. The library inductor can only be an idealised version of the real component. Many physical design aspects such as loss and skin effect factors can not be modelled with any accuracy, if at all. Also, the makers usually supply Q values at only one test frequency. your filter will no doubt be working at some other frequency and the Q will be different. Maybe best to grade the components on a commercial Q meter. 2% ripple is not a lot, and a simple bench Q meter setup (generator, series L, parr C, AC voltmeter, would not easily offer up the required precision.

============================ Its just wL/R isnt it? Or resonate it against a known C, measure hi and lo 3dB down freq points, then compute f0/bw

Yes it is wL/R but the damned R will refuse to be easily pinned down. A simple way is to series resonate the inductor and capacitor. That way there is fewer complications from shunt loading effects dropping the measured Q value. Use a signal generator with a 50ohm output. Feed the output voltage to a