Hi! Depends on your situation.
First, thanks for letting me know that you are getting RCs with only one question—it's not supposed to do that! I've asked our dev team to look into it. Until we can get that fixed, don't include RC in the Random Generator, but then pull up an RC passage yourself and do 3-4 of the problems at the end of your Random timed set. (And then enter the answers in Nav manually.)
(One note: It's supposed to suppress RC entirely if there aren't enough to fulfill the request—eg, if you ask for only 1-2 problems in a set or if you ask for 3 problems but you've already finished all of the 3-problem RCs in the book. So just make sure that you're asking for at least 5 problems when you include RC in the mix. If you're already doing this...then do what I said above until our dev team can figure out what's going on and fix it!)
Ok back to how to use OG:
(1) If you are following any of our study plans that include OG assignments (any of our live courses, Interact, the Self-Study Toolkit), then follow the OG assignments as given throughout the syllabus. Do log your answers in Navigator. When you get to the end of your syllabus, you can then use the Random Set Generator to do additional problems that you didn't already do during the course (Nav will know which ones you already logged).
(2) If you aren't using any of those study plans, then I would use the OG in these ways (depending on the stage of your studies):
--> Earlier in your studies: As you read / study each new topic or question type, do a few OG problems of just that type. Eg, if you just studied quadratics or modifiers (sentence correction), do a couple of each to solidify your skills (not all—save some for review later!). Keep a list of any that you want to try again later (and mark them to Redo in Navigator).
--> Mid-studies: When you're maybe halfway through your initial learning phase of each question type and topic area, start to include a problem or two from older topics—just to keep things fresh. Some may be new, but also redo the ones that you marked at the prior step. (Continue to mark any that you want to try again in future.)
--> Later studies / review period: Once you've made it through your initial learning phase of each question type and topic area, now you're ready to test yourself using test conditions. At this point, do random sets that mimic the real test—ie, you could get anything. For quant, choose DS and PS in the Random Set Generator, but don't choose any topics (so that it will include all topic areas).
The instructions pop-up within Navigator also talks about all of this and gives recommendations (including what difficulty buckets to choose based on your score on your last practice test), so if you haven't already read through that, please do so.
And let us know any feedback you have! We'd love to make Navigator even better.