Nigel, I'd be surprised if there were many (if any) instances where an engine failure or similar (belt fail on a Robbie perhaps) at low level has been the cause of an accident (fatal or not). Most low-level accidents are attributable to CFIT or, more usually, wires.
However, ignoring the H-V curve is fraught with danger and pilots must be aware of the extra risks involved when operating inside it. Chopjock's assertion that the flare will give you time to sort things out ignores the reality of the time it takes to recognise the failure and take corrective action, especially at speeds where there is little flare effect - say 60 kts and below - to save your backside.
Your best option for a succesful EOL is to be into wind, wings level, at your recommended autorotation speed with the minimum RoD possible and over a flat and level piece of ground suitable for the manoeuvre (or in the low hover). That doesn't mean to say you won't survive if those parameters aren't met but the further you are from them, the less and less likely even a gifted pilot is to be able to walk away.
If you are habitually operating outside those conditions then you would be very stupid not to recognise that you have ratcheted up the risk significantly both from the position and the extended exposure. Again, this isn't a problem if you are cognisant of the risk and try to mitigate it where possible.
The German Air Force used to teach pilots to raise the lever before flaring in the Huey in the event of an engine failure at very low level (NOE) to avoid smashing the tail in and making things worse. But that was in a very high inertia rotor system and the same technique in, say, an R22 wouldn't be appropriate as you would be counting the Nr in single figures very quickly.
Bottom line - the H-V curve is there for good reason - your safety
