The following are the distibutions (created iteratively for a sample of 10,000 rolls for each row) for three methods of generating successes. The first is the orthodox Tabletop method (roll D10s, successes on 8 or more, etc). The second is the MET single card draw with a boundary of 5. Lastly is the "Three Card Draw" method I concocted, which seems to work better and quicker than the standard MET system; further details are here.
NB: I've yet to find a satisfactory method to replace the Chance Die in the Three-Card Draw system, but as it doesn't affect the working of most tests I've left it out.