Comment by chrisdotcode

So in 2019, you're saying that LE would have been fine with serving 50% of all Android users certificate errors, is that correct?

If so, what would make them suggest such a plan in the first place?