It's not an argument, it's a worthwhile discussion as pretty much every misconception has come up so far.
I'll start with a t-bone accident. If they've clouted you side on, then it's either a SMIDSY as they looked at the thin profile of the bike then moved in to you, or they were carrying excessive speed in the first place to hit you side on. This kind of accident rarely happens outside a SMIDSY occurrence.
What makes a bike imperceptible is the profile, you're thin, and you will blend in to what is behind. Especially if your one headlight is in line with a headlight from a car following you. A bike is thin, and because of that it is hard to judge distance which means it is hard to judge speed. Whatever colour your bike, from the front-on there is NOT enough colour to make a significant difference in the eyes of an observer - WHO IS NOT EXPECTING TO SEE YOU.
And that is the crux of the matter. The human eye can not perceive everything it sees. It can only only see what the main focal area rests on for a short while. In-between is where you see almost nothing other than a vague blur, and unless there is a massive change in brightness or colour, it gets missed. We see smooth movement, only because our brain interpolates the bits in between where our eyes rest.
So back to colour. When does colour stand out? When there are large blocks in contrast with the background. Front on you're not in contrast. A solid colour helmet or a solid colour bike is not as good as a bike or helmet with a bold colour stripe on it. Search the Shoei Notch Borealis helmet, or look at the classic GSXR colour, big blocks of blue and white. All these things stand out. Let's look at the EU standard Hi-Viz vest. Bright yellow, reflective strips for catching light at night, but hang a tired one against a background of fading greenery, it's not as obvious as if it was against a blue background.
The other thing with these statistical surveys is you have to consider the weighting. Who commissioned the survey, why did they commission it, and what was needing to be proved. If you set out to prove that dark bikes are less safe that bright bikes, then you'll find the link to it. But also, how many black bikes are there on the road, how many white ones are there? Was this survey in one area, what's the distribution of bikes in that area?
White bikes and white helmets CAN stand out more than black, and I'd support that in many situations they will stand out more than black. BUT in (say) London where there are a lot of white painted buildings, white will blend.
Personally during my summer riding, I find darker helmets show up more and in winter, white ones do.
There are millions of ways to be seen, and if you can't be seen you can ride defensively to suit the circumstances at that time. If you approach a side turning with an approaching car, one of the ways of mitigating the situation is to move away from the side turning. This firstly gives lateral movement to the car at the side turning, which is more noticeable, and it also moves you further from the danger giving you and the driver more time to react should the situation develop. However, some of these are advanced techniques and moving wildly across the road should only be done if you know WHY you are doing it and if you UNDERSTAND what you will achieve and that you are not putting yourself in to other danger.
There is no blanket answer to what is the best colour or what the most dangerous colour is, it's hugely dependent on the circumstances.