huge discrepancy in G-PREP software scores in the last 4 exams.

Some variance is normal in GMAT scores (and on the scores of any multiple choice test, adaptive or not), because on some questions you'll have guessed, and on some tests those guesses will work out well, and on others they'll work out unluckily badly. The scores you've posted above wouldn't be completely unusual for a test taker who was around the Q45/V35 level, which means you're close to your target right now. Because variance can sometimes work to your advantage, that means you'd definitely have a chance to hit your target if you took a test today (as you proved when you got that 700 score), though if your level is slightly below your target, that chance will be less than 1/2. You'll improve your chances a lot if you can make even some small genuine improvement before test day.

You can get two slightly different overall scores with the same scaled scores because scaled scores are rounded off to the nearest integer, but the overall score is computed before that rounding is carried out. So if you have a Q44 / V33, that could actually be more like a Q43.6 / V32.6, or could be more like a Q44.4 / V33.4. Those two combinations might produce a slightly different total score out of 800.

Based on your debrief, it does sound like a good idea to do the Verbal section first. If that's your plan, definitely do another mock where you use the section order you intend to use on test day, to confirm that it's a good idea (i.e. confirm your score doesn't drop), and also to become comfortable with what you'll experience on the real test.

Good luck!