So, this is fun. France has a new official keyboard layout: https://norme-azerty.fr. There’s also a nice announcement in English, linked from that page.
That news by itself would probably be of interest only to French speakers (French exceptionalism strikes again!), but the approach seems of interest here. The researchers built a set of analysis tools and a corpus in order to help the French government optimize a new keyboard layout that would make typing easier, without causing too much disruption to muscle memory. The letter characters all stay in the same place as AZERTY, but the symbols and accents move around to reflect French usage.
Even though the main project page is in French, a lot of the research was published in English. Here, for example, is a paper titled Observations on Typing from 136 Million Keystrokes Of particular interest here is their quantification of the relation of typing speed and rollover:
While slow typists use almost no rollover, the majority of fast typists use rollover for 40–70% of keypresses. Trained typists average ∼5% more rollover than untrained typists. When rollover is used, keystrokes overlap by 30 ms, on average, and up to 100 ms. The prevalence of this behaviour was surprising.
Earlier in the paper they point out that faster typists also make fewer errors. I’m not a fast enough typist for it to matter (one semester of typing class in high school got me to around 70 WPM but I have settled comfortably around 55 in the years since, making me an average typist in their sample set) but I would guess that “correct” interpretation of rollover would be a requirement of a fast enough typist. That way the rollover doesn’t induce errors not present in the typing itself.
You can still take the test but I don’t know that it will add anything to their research.