The cycle count starts any time there is a disruption in the dose. In other words you start counting after an increase, decrease, skipped shot, fur shot, giving a slightly reduced dose, etc.
When you increased the dose from 1.0u to 1.25u your cycle count started over due to a change in dose. As of today, you are at 17 cycles at your current dose. The cycle count did not restart at any time after 2/7.
I think the point that Wendy was trying to make is that when a cat starts to see an improvement in nadirs -- nadirs are in the <200 range, you hold the dose for 8 - 10 cycles. You don't restart the cycle count -- you tack on the extra days to what is usually a 3 day cycle count. However, Zoot is not new to blues and greens so that point doesn't apply.
The basic rule of thumb is that if nadirs are > 200, hold the dose for 3 days/6 cycles before increasing. If nadirs are < 100, hold the dose for 5 days/10 cycles. This doesn't mean only one nadir -- most of the nadirs need to be in the range noted.
Ultimately, with TR, you will want to shoot any number above 50. I realize that it's counterintuitive that shooting a number in the 50s won't cause a hypoglycemic episode. A Lantus curve is flat. Take a look at
Pax's spreadsheet. You'll see what happens with shooting low numbers -- the curve flattens out and ideally, the kitty stays in really good numbers.
Does this help?