Motorcycle Action Group got involved in this when they were trying to legislate motorbikes to have permanent lighting. The main proponents trying to drive this thru' were the car manufacturers, as it reduces their manufacturing costs.
The quality of claimed research and conclusions drawn from it were a bit suspect. Especially the USA one quoted above, seems to have been funded by the motor industry...
The US data on drivers and vehicles is not generally applicable to the rest of the world, although they would like it to be. They are basically crap drivers, fall asleep or drink/drive/drugs more. The US "safety" research is being used to justify autonomous cars in the UK/EU, there is no such data or evidence in the UK or EU to support this.
The Australian study on all permanent lights on research highlighted this, not that there was no accident reduction, but that it is much lower than most of the "safety" folk want to claim.
They also tried to apply car driving stats to motorbikes, which simply does not work. To date I'm not aware of any valid research post-legislation to valdidate the claimed reductions, or on motorbike safety with lights on.
I ride both older bikes with light switches and new without. I've had more grief with the lights on from annoyed drivers who think I am flashing them or it blinds them when behind them (as the headlight is higher than on a car), and no noticeable benefit or difference from riding with light off. Basically if the driver is not paying attention nothing make a difference. Ditto wearing hi-vis, although in peak rush hour hi vis seems to get them to back-off a little bit when following behind me.