You can only use official tests to accurately assess yourself, and to evaluate how well you're pacing yourself. The tests from any prep company (it doesn't matter which) will have built-in biases -- company questions will overemphasize some skills and underemphasize others. So you might do better or worse on a company test than on the real thing, depending on whether you're good at those things the company overemphasizes (and of course the scoring of a company test might not be realistic either). There's also no reason to think company questions will be realistically time-consuming, so you can easily reach incorrect conclusions about how successfully you're pacing yourself, and thus make pacing adjustments you should not be making, if you rely too much on company tests. You can probably see this is true just from your experience, if you compare your company diagnostic test scores with your official one; they're not even remotely close. I've found that's true for many of my students as well; some have similar scores on company and official tests, and some have wildly different scores, on those same company tests.
It's not a good idea to use diagnostic tests to assess how weak or strong you are at individual topics. The sample sizes tend to be far too small, and you need to account for question difficulty, and for how lucky you were if you needed to guess. If you had, say, four Geometry questions on a test, and answered two correctly, that would not be a good performance if the four questions were all easy. But it wouldn't be bad at all if the four questions were very hard. You'll get much more reliable information if you evaluate yourself using larger pools of official questions; if you solve a couple dozen official questions scattered throughout one of the official books (so the difficulty level spans the full range), and do the same for other topics, you'll get much better information than you could from a diagnostic test about what you're weakest and strongest at.
Diagnostic tests are great tools to practice pacing, but when you do that, you're really practicing making time investment decisions. You should be evaluating, perhaps a minute into a question, whether it's a good idea to invest more time in the question, or whether it would be better to take your best guess and move on. In math, it's usually true that if you don't see how to solve a question quickly, you might not see how for hours, so it's important to be disciplined about moving on from the hardest questions on the test (getting those wrong really doesn't hurt much anyway). And when you evaluate your performance on any test, you should look back over those questions you spent longest on, where you did decide to invest a lot of time, to see how well you are making those time investment decisions. If you have trouble recalling, after a test, when you did that, you might just take some very brief notes during the test about where you're spending time, and about where you're quickly guessing.
Your official diagnostic score is excellent, so it sounds like you're in a great position to reach your target (you might already be where you need to be, but you'd want another official test score to reliably gauge your current level). Given your time frame, I'd strongly suggest you use only official practice tests leading up to the real one, since they'll simulate the test you're going to take much more closely than any other test could. Good luck!