I have some bad news, Marcab: GMATPrep 1.0 and GMATPrep 2.0 contain exactly the same questions. The GMATPrep 2.0 software looks nicer, but it seems to use the same question bank for both test 1 and test 2, so you're not even guaranteed to get fresh questions when you take the second test. If you have the choice, I would recommend using the older software for quant and verbal, since the question banks are distinct for the two tests within that program. The only problem is that the old software doesn't contain the Integrated Reasoning section, but I suppose you could substitute with questions from some other resource...?
Maybe somebody from
Manhattan GMAT can chime in on this, but I'm pretty sure that they do extensive testing on their own algorithm, and test takers will, on average, earn roughly the same score on the
MGMAT tests as on the actual GMAT. But everybody is different, and I'm pretty convinced that GMAT students who intensively study
MGMAT materials will do disproportionately well on
MGMAT exams, since they're accustomed to
MGMAT's question-writing style. On the other hand,
MGMAT's quant questions are wordier and more difficult than actual GMAT questions, and some test-takers get really, really stubborn on
MGMAT tests, and their scores suffer as a result, even though the
MGMAT algorithm seems to compensate for the increased difficulty of their questions.
So I guess the bottom line is that
MGMAT tests are pretty darned good, but any individual GMAT student might do better or worse on them, depending on their background and test-taking habits.
I would argue that no "unofficial" test is ever going to be a perfect copy of the actual exam, since the GMAT questions and scoring system are so incredibly nuanced.
MGMAT does an an amazing job, but their scores may or may not be accurate for any individual test-taker. So do your best to learn from every practice test, but take the scores with a grain of salt, unless you're using completely fresh GMATPrep tests.
I hope this helps!