You are absolutely right, that (for plain, domain attestation) paid CAs are exactly as trustworthy as LetsEncrypt, and often much less (remember the Diginotar d

You are absolutely right, that (for plain, domain attestation) paid CAs are exactly as trustworthy as LetsEncrypt, and often much less (remember the Diginotar debacle, for example). "Keeping the streets safe" is not their responsibility, except in a very limited sense. The $500 extended validation was mostly paper work and snake oil. My point wasn't to discredit LetsEncrypt, but to point out that Google's claim to mitigate the MITM attack vector with https-first wasn't a very strong argument. I mean, yes, sure: if you can't intercept or downgrade to HTTP the MITM doesn't work. But all the HTTP seems to do was redirect to a malicious payload. But you can also do a redirect in HTTPS. So if you can spoof someone to go to https://g00gle.com/ https://g00gle.com/ it should be just as easy to launch the attack chain from there.