Instead of 20,000 little nozzles, why not just use one big thruster? it's likely to use a lot less material having a lot less external surface area and therefore weigh a lot less, increasing the power to weight ratio - which is far more important than how slowly you can waste, I mean burn the fuel by spending longer on the "wrong" side of the accellerational curve.
Those figures you quote for arriving at, say Mars, in 15 days - Lets ignore the fact that this orbital trajectory wouldn't leave any possibility to hit the Earth if you miss Mars.... You've taken into account decelleration? By wasting 2/3'rds of the fuel burning it so slowly, it'll weigh a lot less by the time you have to think about slowing down - But you'll still need to be slowing down about or before the ½ way mark. Will you actually have enough fuel to be able to slow down?
If this was such a good idea - why couldn't we just print these nozzels ourselves?
Again, what we really need is propulsion technologies that don't use propellant. Propellant based tech has hardly improved really since the appolo missions, and the only thing being built to give out anywhere near as much thrust as the Saturn V5 being NASA's "new" SLS. Technologies like the EM-Drive - though in their infancy and thusly largely ineffective - could output more than 10 netwons of thrust with 10KW of electrical input. No propellant. It's quite simple to generate 10KW too. Don't need a fuel tank or several hundred thousand kilos of fuel, or have to try and fight gravity with them - and can thusly maintain a more constant power/weight ratio and further, constantly apply power opening up long distance and long duration missions.
