Because, usually the kind of people that’d opt in are techies/power users or work (volunteer, or paid) for Mozilla themselves. Let’s say only 1% of your userbase opts in to this, how is that not biased? (As it currently stands, I believe this is a further optin in a tucked away menu).
Then they should make the possibility to opt-in more prominent, instead of switching to opt-out with the option tucked away in some menu where only techies/power users will disable it.
The Mozilla user base is already biased. Plus they can run experimental test fixtures to canvas sites. This sounds like a solution looking for a problem.
>They claim is biased but is it really biased? How do they know? I think this is just making up excuses so they can collect more data.
Informed, constructive opinion there.
One clear sign of the bias is that the crash rate of the browser goes up massively every time a new version transforms from beta to release. Clearly, it's not renaming that string that makes the browser crash. The populations are just fundamentally different.
To give an obvious example, beta users are overwhelmingly more like to have up to date video drivers. (Which can be seen in crash reports, but is also very logical).
They get good enough data from the people that have volunteered it. I don't know what makes them think it's biased but I seriously doubt that is true.