Center tapping
OK - So as far as I can tell, the center tapping is done here to allow them to split up the tuning capacitance into an MSB part and an LSB part. If we have a 2:1 stepdown transformer (this is equivalent to their center-tapped inductor), then impedances on the secondary side must be 4 times larger (2*2) to have the same effect as if they were on the primary side. So if you want to have some tuning capacitance on the secondary side (i.e. connected to the center turn), its effect on the tank resonance gets divided by 4 -- this allows them to implement a "fine" adjust, which combined with the MSB tuning capacitance across the entire inductor, gives them a good tuning range.
As far as the Q goes, I really don't think the center tapping does anything. The effect of resistance in the LSB capacitance is magnified by 4 times when you move it to the primary side, so unless you can reduce the parasitic resistance by more than a factor of 4 for a 4x increase in capacitance, it's not likely to yield any benefit.
Where does the high inductor Q come from then? I believe it is described on page 3 of the paper - They maximized inductance for a given area to reduce substrate losses, used SOI, and etched all the way down to the insulator. That means there is very little loss from the substrate -- no eddy currents, etc...
As far as the varactor Q goes, they're not using typical varactors -- they use constant caps with voltage controlled resistors -- see page 2 and reference 8.
Hope this helps