There was already a lot of debate about adding the CA, but the final consensus was, I believe, "as long as they don't actually abuse it, we should trust them; as soon as we have evidence of someone doing a MITM attack with that CA, we will blacklist it".