I have done the spreadsheet exactly as Coppock did it in his paper - he showed a spreadsheet that he used to do by hand in the days before computers. I suspect that he divided by 10 (just move the decimal place) to save a step in the calculations and also to keep the numbers manageable, since he also had to do the plots on the graph paper by hand in those days, otherwise he might have omitted this step altogether.
You are correct that the sum of all the weighted values should be divided by 55. If it was the individual values that were relevant in absolute terms, then that is the only acceptable method. However, the Coppock indicator only seeks to pick when the line turns up below zero. The divisor does not determine the sign. Also, no matter what the divisor was, the shape of the curve would be the same for the purpose, would it not? So, I left it as Coppock did it.