The serving size was assumed to be the maker's
On this page11
The recommendation was treated as the serving
It is not. The regulation determines serving size by reference to 21 CFR 101.9(b) and the reference amounts in 101.12(b) Table 2.
Which is the same table the edibles property uses
Reference Amounts Customarily Consumed, where all other candies are 30 grams and syrups are 30 millilitres. Being labeled a supplement does not move a product out of it.
So servings per container is not a free number
It is the fill divided by the serving, and only one of those two is the maker's. Choosing the container chooses the declaration.
And a tidy count was designed backwards
60 servings is a commercial shape, and it has to fall out of the arithmetic rather than being asserted on the label.
The fill mass was estimated rather than weighed
A gummy's mass per piece is a property of the recipe. One filled jar on a scale settles it, and we publish no fill weight because any figure we gave would be invented.
A volume was used where the table asks for mass
Which is the same gap the edibles property states: our capacity field is in millilitres and the candy reference amount is in grams.
The piece size was changed after the container was ordered
A larger gummy means fewer per jar, which changes the declared servings per container and therefore the label.
Three leaves were treated as one product
A gummy, a chew and a chewable tablet pack at different densities into the same jar, so a fill weight from one does not transfer to another.
And the liner was left out
49 of the 255 closures here have no liner or no record of one, and a gummy in an unlined jar fuses. That failure looks like an old product.
Weigh first, then declare, then order
Three steps in that order. The container is the least reversible of them, so it goes last.