Intersting question on the overspeed. I agree with LZ on what would happen if the governor were switched on during overspeed with govnr out initially. It has the ability to add and remove throttle input and is constantly doing so, eg. when flaring, when a gust of incoming airflow enters the disc, when the pilot uses right pedal in the hover etc. It would react just the same in the above conditions.
I'm not so convinced that it will raise the lever slightly to control excess RRPM as I understood the latest (Mk 3) gvnr did not move the lever like the Mk 2 model used to (lots !)but I'd be corrected.
An intersting point is that if the governor can only increase and reduce throttle input, it is only controlling RRPM through that. It doesn't have any mechanism (a brake or similar) to slow RRPM if they were increased above the range of governor/throttle control by an aerodynamic effect on the disc and I wonder that if some flying technique or manooeuvre caused a huge increase in RRPM it may be out of the governor's authority range and it would not be able to reduce RRPM.(If it does still increase collective pitch by rasing the lever then the increase in drag would reduce RRPM). So, in theory at least, the upper limit of governor operation has to be when the governor can no longer reduce RRPM by reducing throttle.
Incidentally, I had an R22 governor fail a few years ago which allowed RRPM to rise to around 110% in straight and level flight before it was switched out. The fault was traced to a faulty logic unit/signal generator which is fitted behind the LHS back, on the cockpit side of the firewall.
One of the main concerns (and there are many)
about overspeed is the amount of centrifugal force exerted on the MR blade attachment bearings ('spindle bearings')is 4.5 tons at 104% but will rise exponentially to 9 tons if allowed to reach 114%.
I'll find out about the horn operation in the morning, interesting Q! I'm pretty certain it is a system that blows the horn if impulse rate from the left mag becomes too low, as opposed to being triggered by the needle, but we'll see.