The following should be taken as strictly my opinion. That said, I have extensive experience making carbonated drinks with the bubbles added from a CO2 tank, and almost none (since being a teen back in the 60s) using yeast.
In experimenting with new (to you) flavors, I would highly recommend attacking one variable at a time, noting that the interaction between yeast growth and whatever is used for flavoring is a variable. Whatever is used for sweetening, and its interaction with both yeast and flavorings, are also variables.
My approach to a new flavoring (cinnamon extract for example) would be to make a series of batches (1 liter at a time) with different quantities of extract (or different quantities of two flavorings together if that was applicable), and holding the amount of carbonation, the type of sweetener, and anything else I could think of that might matter, constant. Once I have worked out what I consider the "best" only THEN would I introduce other variables, such as the type of sweetener or its amount.
Of course, if one has the resources and is a good record-keeper, one can use more than one variable at a time. With four batches you can made all combinations of two amounts of flavoring and two amounts of sweetener, but things get complicated very fast if you try to ramp things up to more variables or more levels of a variable.
For myself I would not bother with using yeast for carbonated drinks, but that's just me.
In experimenting with new (to you) flavors, I would highly recommend attacking one variable at a time, noting that the interaction between yeast growth and whatever is used for flavoring is a variable. Whatever is used for sweetening, and its interaction with both yeast and flavorings, are also variables.
My approach to a new flavoring (cinnamon extract for example) would be to make a series of batches (1 liter at a time) with different quantities of extract (or different quantities of two flavorings together if that was applicable), and holding the amount of carbonation, the type of sweetener, and anything else I could think of that might matter, constant. Once I have worked out what I consider the "best" only THEN would I introduce other variables, such as the type of sweetener or its amount.
Of course, if one has the resources and is a good record-keeper, one can use more than one variable at a time. With four batches you can made all combinations of two amounts of flavoring and two amounts of sweetener, but things get complicated very fast if you try to ramp things up to more variables or more levels of a variable.
For myself I would not bother with using yeast for carbonated drinks, but that's just me.
