Earlier quoted context omitted.
This is the important point everyone seems to be missing. If you look at the example tweets they all have some aspect of making the [Religious Group] out to be less than human: rats, virus, maggots, etc. People can still freely express their distaste for any and all religion on twitter so long as they don’t dehumanize the people who follow that religion.
Application will be different from stated intent. How much you want to bet ? haven't we all seen this in practice long enough now to know how this little routine goes?
Rules & regulations should be written, evaluated, and revised the same way code is. Code doesn’t always run as we intended but we (usually) don’t crucify the developer. We learn from the misunderstanding, fix it, and then re-evaluate when the next inevitable misunderstanding comes up. We get feedback and we iterate. That’s exactly what Twitter is doing. Their first policy didn’t work out as planned and this is their next iteration based on the feedback they received on policy v1.0.0
And will Twitter update this policy yet again once they’ve had time to work out some of the kinks? You can bet on it.