Soap is caused by sapponification. Thats water + caustic and oil.
So, always water, but the water can come from different places -
there at the start
there from neutralisation of the FFAs (acid + base > salt + water)
there from mixing the methoxide (im not so sure of the chemistry behind that one)
So, the soap production should correspond to the quantity of 'catalyst' needed (which corresponds to the FFA content, or, indeed, the amount of water present at the start of the process)
You can conclude this by titrating - this will tell you if your catalyst is being used up by simple sapponification, or by the neutralisation reaction, and subsequent sapponification.