My "best" practical solution would be to replace the helical gears with standard ones like DFGT or G25. There will be only a residual axial movement in that case which will be totally handled by the G27's soft spring loaded counter bearings. No more axial pressure on the motor's rods while turning in a specific direction. That's of course just a quick theory.
I personally have a DFGT (very few hours of operation) and I also have the same rattling issue at high frequency FFB spikes, but it's not so prevalent as it is in the G27's and more important it's not unidirectional or vastly pronounced in one direction since there's no angular push from the gears. So I can just assume that the rattling in DFGT's and G25's are there from the residual axial movement, thus I will attempt to open my wheel at some point and try to put some load on the motor's rod, something like the G27 has "by default". I really hope that it will fix it once and for all. It really bothers me a lot.
EDIT: Just opened my DFGT and applied some pressure on the motors rod. No go. There's no axial movement there to be fixed. It seems like the rattling it's simply from the gears themselves having some gap between them maybe a tenth of a mm or even less. Basically there's a "deadzone" between the cogs and at the right frequency and amplitude it goes all harmonic resonance. It even does the rattling while I'm slightly but fast shaking the wheel with my hands. So, no fix for DFGT... And there seems to be and intermediate wheel in the DFGT's chain, while in the G25/27 the motor directly drives the wheel if I'm not mistaken. In DFGT there's the Motor cog wheel > Intermediate cog wheel > Wheel cog wheel...
Yeah, clearly locking the motor's rod in place and slightly shaking the wheel, moves the intermediate wheel for about ~3 tenths of a mm, while the motor cog wheel it's standing still.
So, in the most part, the "deadzone" it's between the small (plastic) motor cog wheel and the intermediate wheel while there seems to be almost no "deadzone" between the intermediate cog and the final wheel cog.
Sooo... I've ended up just greasing the cogs and sticking two pieces of adhesive foam on the full lock ends as there were slamming plastic on plastic at calibration.
Greasing the cogs just slightly helped with the grinding noise about ~15% as a roughly approximation, but now I've learned why Logitech didn't just filled these wheels with grease... Those motors are spinning reeeealy fast... so... they tend to throw grease allover the place... luckily enough, there's enough coverage from the plastic cogs casing to keep most of that grease in one place...
The two pieces of foam really did the trick though. No more slamming on the full lock. Very good damping.
