"Don't Waste your Official GMATPrep Diagnostic Exams" is an expression I hear often. I do think this is true, to an extent (they are the only diagnostic GMAT scores you can trust), but overall I believe that GMAT students focus far too much on them, and not enough on individual question practice / mastering concepts.
Simply put, if you are a high-scoring GMAT student, then full diagnostic exams are not as predictive as you may think.
Why? Because there is a very small difference in performance between a 700 and an 800. Personally, I take the test at least once a year and have always scored somewhere between a 710 and a 770. On the GMATPrep tests, I have scored anywhere from a 690 to a 790. Is this because I did far worse on the 690 than on the 790? No, it's because there is very little margin for error for high GMAT scorers.
As a result, diagnostic exams can only tell us so much. Obsess over them at your own peril! Instead, you should spend plenty of time practicing questions in the OGs, Question Packs,
Exam Packs (untimed), Mobile App, Focus Quizzes, and maybe even the paper-based tests from GMAC.
If you are worried about your ability to concentrate for 4 hours straight, then that's understandable, but you can study for 4 hours straight without having to take a diagnostic exam every time. Breaking lessons into parts is a good thing. Reviewing questions you have done before, and other concepts, is a good thing. Trying a few questions at a time is a good thing. It doesn't always have to be a full practice test, in the same way that not every hour of of soccer practice should be devoted to scrimmages.
Also, remember that
you can take the GMAT up to 5 times a year, and cancel any score that you don't want, without anyone else knowing. Thus, so long as you can afford to re-take it, the GMAT itself has become its own diagnostic test.