Unfortunately I don't have any hands on experience with PV systems, my knowledge is only from one of my former lives as an electrical engineer and inference from theory. My house has trees I don't want to cut down and does not have sufficient north facing roof line for a large system.
However, I did read in the AFR that some country locations are suffering from overvoltage allegedly caused by the farmers who have large PV systems installed (much greater take up of solar in country than latte sipping inner urban types - I drink long black so don't fall into that category!). I would have thought that a PV inverter would have a maximum voltage that it would output, so I'm a little suspicious of the AFR report, but you never know. The inverter would as you suggest have to be in phase with the mains supply, but is there a limit on the voltage they output?
Again, arguing from a theoretical stance, if an inverter is supplying more power than the house is consuming it will be sourcing it to the network. There will be a negligible voltage difference between the internal voltage (that the house will see) as compared to the external network voltage - otherwise there will be a large power dissipation comprising the voltage delta time the current.
I suspect that country (particularly farm house) distribution networks have relatively high losses in the run to the house. In this case there would be a significant voltage drop between a low impedance power source and house, so if the current is flowing back to the network then the extra voltage at the house could reach (say) twice this delta and may be enough to blow household electronic devices (which are often spec'd for US 110/220V rather than our 240V. Of course, this will require that that the PV system output voltage is unconstrained. The cheapest design would assume that the network is such a low impedance sink that the PV system could not alter the voltage. My worldly experience tells me that cheapest and nastiest usually wins out!
