What requires it to be a larger effect? Your claim was that "browser choice" was ineffective because non-Microsoft browsers lost share, but that's nonsense. Microsoft browsers lost share. That there were multiple reasons for this is no evidence that browser choice didn't work. The thing it was supposed to do happened.
I’m quoting from *your source”. How would Microsoft not lose share if they weren’t on mobile? We have a control group - the US. Where there was no browser choice forced on Microsoft, they lost share faster and where Google was more popular.
Losing share faster rather than at the same rate is evidence that there are independent factors involved, which means that it isn't a valid control group.
