Because a CAT test adapts to your test-taking it would be very hard to estimate a score without knowing the difficulty of the questions on the test you took. On an actual CAT you could miss 14 questions and score a 660 and 16 and score a 720 if the level of the 16 questions you missed were significantly harder.

McGraw-Hill, however, being a reputable company probably followed a formula mixing in an appropriate number of easy and difficult questions to approximate a score.

I will use you results and compare them to the paper exam in Kaplan Premier (keep in mind these are different tests, and Kaplan exams tend to underestimate performance). However, if you want a very realistic score to start from, go take one of the tests from mba.com. The GMAT Prep software from mba.com is made from the actual creators of the test and is the closest you can get to estimating your actual score---that should tell you how much to study.

As for my rough current guess:

Q 41 (64th percentile)
V 45 (99th percentile)

660 (90th percentile)

the disparity there is quite large (between sections) so take a practice test.

Although, at first glance, you should seemingly focus on quant.