Just saying that more plates are/are not better is misleading. How are they configured for your given input voltage?
A +NNNNN- configuration at 13.5 volts yields X MMW.
A +NNNNNNNNNNNN- configuration at the same 13.5 volts won't produce much as there is not enough voltage across each plate to generate HHO. Thus a very low MMW.
However A +NNNNN-NNNNN+NNNNN-... yields a better X MMW from the first example.
This assumes that all things are created equal in the HHO generator and just the plate configuration is changing. It is also assuming that all electrodes inside the electrolyte are shielded from said electrolyte. And that the edges of the plates are all sealed. Without any of this your testing is invalid as the stray currents are going to give you random results.
Hope this helps.
--
Some days I get the sinking feeling that Orwell was an optimist!