I think what you are seeing a lot of is "bouncing".
Looking at the AM cycle Diablo had on the 17th - That green number caused a bounce. Not only is there an immediate release of glucose when that happens, which will raise BG pretty quick, but there are also other hormones released. The quick boost of glucose will tend to clear out of the system pretty quick. But sometimes you see lingering effects, and in some cats it can take a couple of days to settle back down. That is what I think was going on all day, and into the next day, the 19th. It could be that he dropped lower than the lowest number you happened to catch on the 19th (146), and the ugly pink and red numbers today are just some residual "bouncing" going on.
All that said, I agree with Robin. You wouldn't necessarily shoot the same dose on a 200 that you would on a 400. But that assumes that all those numbers are not part of a bounce from a low number in the past couple/three days. This is the "hard part"....figuring out if the highs are from a previous "low" BG level, or if they are just due to the dose being too low.
A true sliding scale might be appropriate if you have a good idea of what is causing the high numbers. That's because Prozinc is what I like to call an "if/then" insulin. If you shoot X units, and see a logical drop in BGs, then shooting 2X units should give you a bigger drop. Not necessarily twice as much just because the dose is twice as big, but there's some degree of "logic" to it.
See, the thing about bounces is that they eventually "clear". All that extra effect from the hormones goes away. The danger in thinking "I got a high number, I need to shoot more insulin this time" is that IF the high readings are due to a bounce, and it decides to "clear" after the bigger shot, then all of a sudden, you have more insulin on board than you need, and it can cause a pretty low BG level, and lead to ..... you guessed it!.... another bounce. ohmygod_smile
This is when we look not just at the numbers themselves, but on the "patterns" that show up on your spreadsheet. For instance, does it look like something happens over and over again. Does he drop "too low", and then see 3, 4, or 6 cycles of higher numbers, then you see a nice low AMPS or PMPS that leaves you wondering whether or not to give a shot? And that would be an instance where you would look back and say "wow, maybe I should have kept the dose constant because then I wouldn't have to wonder if the dose amount was confusing things while all the bouncing was going on".
I'm not sure if that helps explain things, or makes things more confusing. This FD thing is like a giant jigsaw puzzle where there's no edge pieces, and all the pieces are various shades of grey. It gets really frustrating at times, and occasionally, all you can do is experiment by trying different things. The keys there are first, "am I going to be able to be around to get tests and collect data?", and second, "what happened the last time I saw the same sort of numbers and patterns?"
Carl