The media plays fast and loose with terminology either out of plain ignorance or the desire to sell a story.
Maybe I'm remembering my QED special topics class incorrectly, but I thought the state teleportation is believed to be instant. The problem is to extract the state information, the transformation operators still need to be communicated. So you still have to get on the phone to tell your colleague across the planet what to do to see the entangled result. It's been years, so maybe I'm a little off on that?
Causality also isn't a hard-and-fast law. We've just never seen it violated before. It's one of those physics problems that is really hard to find evidence to support or discredit it.
I understand the temptation to think that every "law" is just waiting to be broken, but I can easily read the same data and come to the opposite conclusion about the strength of our current "laws".
Why can't it happen simultaneously? Why is distance relevant in this case?
See my top-level post for more information [1].
This is the part where you are potentially wrong. Entanglement is about correlation. In order to derive the original message, you need the local entangled state, along with data from the remote actor. You could then correlate this information to derive the original remote entangled state -- ie, some message. Of course, this is now useless to you, as they could have simply sent the original message and saved you the correlation step.
However, I don't know of any reason that the local quantum state could not be measured prior to the arrival of the data from the remote actor. The math says that the measurable effect should happen instantaneously. But that data alone is not sufficient to derive a message.