I agree with Rodney. The LT head would be my choice. And it also has the advantage of being more "blade friendly" as it allows for the use of a bigger blade selection.
The rigidity of any rotor head, is not just related to the rubber dampers and composition (metal or GRP plastic) of the head components. It is also related to the flex (before bending occurs) of the steel spindles (Note: try bending one and see how far you have to do so before it will bend, and not spring back into its normal position) , the rotor hub/yoke and the flex in the swashplate stars and other control system components
I have tested properly molded GRP blade grips that flexed less then some spindles they were attached to!! Also if you want more collective pop from the Hawk (assuming you are using a good hi-torque digital servo), just reinforce or double-up the collective arm, and you will see a big difference.
As the LT head has a fairly beefy spindle (7mm), you will find it much stiffer then the 6mm one used in the 50 nitros which weigh in at more then a pound less ready to fly. The LT head's CNC rotor hub is also adequately rigid.
These are why Rodney and so many others find the LT head does well in 3D at relatively high rotor speeds.
Where you may want to improve things, to insure a more rigid control system, is to replace the stock swashlate, which has a metal inner star and a GRP outer star (as on the Hawk & Swift 16 & 550) with the all metal one from the Raven 50/NX50/Predator. I think a lot more would be gained from that change.