Yes I think she often nadirs then, so a little earlier than +6
Until they are very well regulated the nadirs do move around, George nadired as early as +2 and as late as +12., though he more often was around +5 +6. He wasn't overly carb sensitive first time round.
I think her onset moves around
They will do
I like testing her at +3 a
I think that by the time you test at +3 she has often already onset an hour or two earlier.
Let's look at today, she started at 168, my guess is that if you had tested her at +1 you would have seen a drop instead of a spike from the food, that doesn't mean that she hasn't had one (or is not carb sensitive) it's just that it's masked by the insulin onsetting early.
Take 11/30 pm cycle, started at 100, by +2 she was 77, so she also had already onset by +2, the cycle started somewhere between +1 and +2, or maybe before +1.
Take 11/29 pmps122 +2 110, +2 almost the same as pmps, no food bump, again this was an active cycle, so that's why you didn't see a food bump, on this occasion, if I had to guess I would say she onset closer to +2 than the previous two examples and if you had got the +1, you might have seen a slight rise in BG.
I think if you started regularly getting +1 and +2 you would soon see that her onset is often in that area.
On the day she had dropped low by +3 and it took a while to get her to stay up, it is likely she would have signalled that to you with a sharp dip. maybe into the high 50's or low 60's. How does that help, well it helps because you can feed at that point, and you can target the carbs. For instance, if I saw a sharp dip at +1, I would give a higher LC instead of his usual 3% ish he would get an 8 or 9%, maybe higher if I couldn't monitor for some reason.
With George a drop at +1 often signaled a very active cycle, and because he wasn't carb sensitive I had to get proactive with the carb choices. If +1 was about the same then he would probably have a normalish cycle, +2 would tell me more.
So my advice would be start testing +1 /+2, depending on what the numbers suggest then you could skip some of the latter tests. Move the testing around a bit more, once you get a feel for it.