When i look at her spreadsheet, I don't actually see bouncing. I see the insulin getting used up before the end of the cycle, and as it runs out, the numbers rise. That's different than bouncing. In bouncing you usually see a drop and steep rise (your spreadsheet shows more of a drop and then a surf - 6/19AM and 6/26AM, or a drop and then a gentle rise 6/21AM and 6/22AM). In bouncing, the drops are usually followed by a few high flat cycles with very little insulin response. You don't have those high flat cycles. You have good cycle after good cycle. Even your PM cycles where we don't have a lot of data, do have enough to indicate that she's responding to the insulin. And again, the morning cycles are all showing nice responses - not high flats.
So yes, I would have kept at 2u, and maybe even done some very small sneaky increases from there to try to lengthen the duration of the insulin without sending her diving.
As for food, it can impact things, but I tend to think of timing the food as being more of a factor as they get better regulated, and I tend not to worry about it too much before then. However, I know over on the Lantus forum, they do a lot more with using food to manipulate the curve, so I would definitely defer to Linda on that. Just not an area where I've seen as much success and wouldn't know how to guide you on that.
As for the amount of food, she's only 5.6 pounds, so that sounds like a lot of food, but my hunch is that she's eating that much because she's still spending time in such high numbers, and probably needs the extra calories. As long as she isn't overweight for her size, I wouldn't worry about the amount. Keep an eye on her weight though in case she starts gaining too much.
Over 360 points lower than PMPS. Just now I got today's +4 and she was 123, 340 points lower than AMPS! I'm happy to see some blues but they're almost scary considering how much and fast she drops.
