Apple is defining unlock in that question as "Has Apple broken encryption or brute forced a PIN for law enforcement in the past?". Apple is stating that no they never have.
If that's true, then how did they get data from phones for law enforcement in the past?
I believe I have a basic understanding of the technology here, so I understand Apple's position. But let me play devil's advocate for a moment so that I can make my issue with their answer on their Q&A page more clear.
The thing that people want to know with that question is, "What's the big deal? You've helped law enforcement crack into the data on phones before, so why is this different than what you already did before?"
For me, reading their Q&A page, it's just saying, "We have never cracked a phone for the fuzz, except all those times in the past when we did". They're not making the answer to that question clear.
Tom, what I think you're saying is:
They've never cracked the encryption on a phone or changed the OS to allow brute force attacks on the password before. In the past, the data on their phones wasn't actually protected very well, so they were able to retrieve the data for the fuzz without the need for those things.
If that's true, then that's what they should say on the Q&A page. But they're not coming right out and saying that in their answer. Probably because it makes one wonder about the chronology:
1. In the olden days, phones weren't secure or encrypted, so the fuzz could get whatever they wanted, right off the phone.
2. Recently, mfrs started securing phones, but they weren't actually very secure, and the fuzz could always enlist the mfr's help to get the data. Apple always complied in those cases.
3. Very recently, Apple decided to really TRULY secure phones, and are now balking at the fuzz's request to crack it.
Looking at the chronology that way, I think this really cuts to the heart of the issue, which is: Back doors have always existed. Apple has recently closed one, and this time, they're drawing a line in the sand at that point. The question is, why is the line at #3 worse than the one at #2? I understand the difference from a technological point of view, but what about the general principle? Apple seemed happy to help the fuzz at #2.
Maybe they weren't happy at all, actually. It looks like the newest encryption features were a time bomb waiting to go off. And I think they knew it would come to this, even as they were developing the newest encryption features. It almost sounds like the development of the new encryption was a calculated passive-aggressive move on their part, a type of protest against the government requiring companies to reveal customer's data all along.
Maybe Apple thought that, after they created these new encryption features, that they could honestly answer the government with "nope, sorry, no way to crack it, not possible". And now the government got clever and said, "but wait, you can crack it *this* way", and Apple was like, "woops".