Q of 10 means you'll probably need a decent RF geared CMOS technology with good BEOL stack... otherwise it'll be tough to get the performance

BEOL (back-end-of-line) is just metal stack

sorry for the jargon...

for good inductors you'll need low resistance metal (therefore thick) which is sufficiently elevated from substrate...

Hi,
Just to continue the line of thought of loose-electron I've seen a type of filters called "Q-enhancement Bandpass Filters" that are proposed as integrated RF Bandpass filters, first time I've seen them was in a tutorial from ISCAS 2004:"Design of Continuous-Time Filters from 0.1Hz to 2.0GHz",Edgar Sanchez-Sinencio and Jose Silva-Martinez, surfing the web you could find some IEEE papers and I've found also a Thesis:"A Q-ENHANCED 3.6 GHZ TUNABLE CMOS BANDPASS FILTER FOR WIDEBAND WIRELESS APPLICATIONS",JIANDONG GE,University of Saskatchewan,Canada.
I suppose that they could be used if power consumption and/or the trade off between NF and linearity are too stringent.

Instead of GmC you could think of using a LRC tank where the tank is somehow related to the VCO tank, then you don't need an extra PLL.

Or you could come up with a smart way to use the RF PLL to tune the tank.

The real problem with both GmC or LRC with Q-enhancement is the inevitable degradation in dynamic range due to using active components in the filter structure (added noise and nonlinear distortion).

There are numerous papers on LRC Q-enhancement, and a few of them deal with analyzing the dynamic range issues. Usually the IMD and noise performance is pretty bad.

But it is cool to play around with for sure.