What to measure instead, if anything
Some readers can't take the break without something to replace the number, and it's fair to ask what. The honest first answer is that you may not need a replacement. The number was answering a question you asked every morning, and the question, as chapters 3 to 8 argued, wasn't one the number could answer. Drop the question for a month and see whether you miss it or only miss the ritual.
If you do want something, here are the candidates, each with its limit.
Clothes. A particular pair of trousers, done up, is a crude measure of the waist that doesn't swing with glycogen the way the scale does. It can tell you over months if something's changing. The limit is chapter 22's case. Trousers can become the number, tried on every morning, with a swing attached, and if you find yourself doing that you've moved the habit rather than dropped it.
The waist itself. The NHS, on its page about overweight and obesity, says that BMI can't tell muscle from fat and suggests the waist to height ratio alongside it, with the aim of keeping the waist to less than half your height. It's a once-a-month measurement at most, and the NHS page shows how to take it. Its virtue is that it's slow and dull. Nobody feels a swing from a tape measure they use twelve times a year.
How you are. Stairs, the walk to the station, sleep, energy. These are what the weight was supposed to be a proxy for, and they're available directly. The limit is that they're vague, and vague is what you were trying to escape. But vague and honest may be better for a while than precise and meaningless.
Nothing. Which is where I'd start. For a month, measure nothing, and let the body do what it's going to do without a witness. If at the end of the month you find you want a measure, pick the slowest one on this list.
What you shouldn't do is swap the daily number for a daily anything. Daily is the problem, not the instrument. A daily tape measure, a daily photograph, a daily trying-on, would each reproduce the swing within a fortnight, because each gives you a point to react to, and you've spent years learning to react to points. The whole of this book has been an argument for lines. Anything you measure instead should be measured slowly enough to draw one.