Hey Sam, my thoughts...
1. 8 valid objections would be rare, so it's possible (as Tom mentioned) that some may have actually been invalid. That would have made the integration easier. But without specifics, I can't say. Although my gut tells me this is likely where the complication originates.
2. It's also possible this was an invalid tension to process (see 3.2.2). Meaning, the proposer was making a proposal based on an "imagined issue" rather than a real example or real felt sense of a problem. Again, not sure, but that is something to look into.
3. Assuming 8 valid objections, then yes, it's quite possible that the first resolution (P + O1), could be "undone" by the next resolution (P+O2), or the next one, etc. The constitution's guidance is simply, "integrate until the objections are resolved," but the process for how to do that is unspecified. Again, hard to say how to handle this in a meeting w/o specifics, but my guess would be 1) try to do them one-at-a-time, checking in w/ each previous objector as well as the proposer to ensure the latest changes are still resolving all previous issues (as long as they are the same issues/objections, meaning they aren't new objections); or 2) resolve them one-at-a-time, but don't worry about "undoing" resolution to previous objections, then once you've completed integrating the last objection, do the second objection round and see how many pass the validity criteria now. Since the constitution says you need to integrate all objections, it may be a stretch in interpretation to say that you can do this, but again, without specifics, it's an option I might consider; 3) given the large number of objections, actually spend some time consciously brainstorming solutions which could resolve multiple objections. Usually not good practice, but in a situation like this, it might be the best way.
4. This is a bad case for asynchronous governance (i.e. email or GlassFrog), which should only be used for proposals which one can reasonably assume, "no one should have any questions about this," and if someone does, then the proposal should be escalated to a meeting. So, glad you went with option #2! )
5. However, even the small working group has a limitation, which is that the bias will likely be toward consensus rather than, "there are no objections," which is a subtle but important difference. Not to mention the other objection criteria (i.e. created by the proposal, known now, and from my role), which could easily get lost in the fray of an unfacilitated "working group" making the whole exercise more complicated, not less. Instead, I would continue to use the elected facilitator and the objection criteria, meaning...I would continue to use the governance process. Instead of a working group, just schedule a special governance meeting for this one issue, and keep scheduling them as necessary until it gets resolved. Not very elegant of course, but some issues are complex by nature, and they just take some time to untangle.
Hope this helps!