OK, low G is a bit of a variable term ranging from 0.99 G to zero G and the answer to recovery depends on the severity (ie the closer to zero) you are.
At zero G in a teetering head helicopter, you have no control power (ie how much you have to move the cyclic to affect the attitude of the fuselage). You can wave the cyclic around as much as you like but it won't affect the attitude of the fuselage.
At zero G you have no cyclic control power so therefore aft cyclic won't be effective and can make things worse (ie chopping off the tail) - the only thing that will save you is to restore positive G and that would have to be with the collective.
At 0.99 G you have a very small reduction (probably imperceptible) in control power so aft cyclic will immediately (if you have any forward speed) restore normal operation.
There is a big sliding scale between those bookends of performance - the lower the G, the lower the control power and somewhere (I don't know exactly) there will come a point where the control power is sufficiently low to render aft cyclic ineffective in restoring G (and therefore control power).
You don't have a G meter in the cockpit so you can only use seat of the pants to assess any apparent reduction in G. You can feel quite clearly a change from 1.0 to 0.5 G and zero is very floaty indeed.
It seems that Bells cope better with reduced G than Robinsons - it may have a lot to do with rotor inertia, or the extra stability provided by the additional features of the rotor head that aren't on the robbies.
Whatever the reason, any reduction in positive G should be avoided (whether through turbulence, exceeding VNE, poor entry to auto) in a Robbie and zero G should be avoided at all costs in any teetering head helicopter.
Soave pilot - I would have to do a search on the AAIB but that is what I remember (it was about 20 years ago) if you lower the lever quickly and push forward with the cyclic (the sort of reaction a FW pilot converting to RW might have to a stall - I have seen this with FW friends) - you will get a reduction in G - how much depends on the rate of application and displacement of the controls.
If you pull up and then put in the same control inputs, you can easily get to zero G ask any military pilot who has bunted a helo in order to scare the troops in the back.
Lonewolf - if you hit a big updraft in the cruise, the natural reaction is to lower the lever to maintain height - if that updraft has caused the nose to pitch up slightly you may well inadvertently push forward on the cyclic as well.