There are 36 cells in series and they are silicon technology, so that would indicate a battery voltage of 0.6 x 36 = 21.6v would be needed in order to pass any significant forward current through them. I tested up to 20v and got no more than a few milliamps; perhaps there would have been a sudden increase if I had tested to a higher voltage. With a simple direct 12v system, there is no need for a diode at all as long as there is a sufficient number of cells in series.
If you start installing a control unit to optimise the energy transfer or to control the charging rate, then there is a definite need for a diode to prevent the control circuitry from loading the battery during darkness and to protect the control elements from reverse voltages. I also take the point from the other contributors: the diode which was already installed was probably to give continuity for series-connected panels.
I've tested the panel this morning to see how it responds to dull, murky overcast conditions. Laid horizontally inside the van above the dashboard, parked facing away from the sun direction and in an allyway between two two-storey buildings, it gave 16.5v off-load. That means that for any reasonable level of daylight, it definitely isn't going to be draining the battery and may possibly be contributing a small charging current.