Only way to know for sure if the lighter colored batches are that way because they contain fewer PPM due to plate out on cathode OR if they are lighter because they are not fully reduced is to try the simple experiment I suggested earlier. Only takes a few minutes. I like to use 2 identical wine glasses. Fill them each to the same level and put a tiny drop of Karo in one of them. Wait a day and you should be able to tell.
Thats definitely a good test to do. However, there are other factors which make one batch lighter or darker than another. Uniformity of particle size (monodispersity) makes a big difference, and contaminants also make a difference.
I have found that the difference between vigorous stirring and less stirring makes a difference. Without stirring, the solutions is not mixed well (homogenous) leading to small particles of silver oxide precipitate. These particles can be small enough that they don't significantly impact the turbidity, but they don't exhibit silver's plasmon resonance and so only appear dark.
This is also where the size of the anode vs current makes a difference, as a larger anode spreads the silver oxide over a larger area and volume of water reducing the bad effect while a lower current gives more time for the silver oxide to diffuse into the water.
Making colloidal silver is a lot more complex than it first appears. But if the 'rules of thumb'
1 are followed, it comes out pretty well.
1) In medieval times, men were allow to beat their wives with a strap as long as it wasn't wider then their thumb. Hence the rule of thumb.