You realize this argument is pointless, right? Since there is currently no such thing as a "positron fueled ion thruster", we can't even debate hard figures for thrust and Isp. You could easily say that the Isp would be 5,000 as I could say it'd be 1200; and you could easily say the thrust would be 10 kN as I could say it'd be .1 kN. It's pointless to haggle over the design points of hypothetical technology...
As for EXISTING electrical propulsion... there's a reason it's used on satellites for attitude control and why DS1 was a
technology demonstrator for it to be used on deep space vessels - because the advancements and improvements to efficiency and thrust for it to be a viable deep space option are... well let's say "significant" to be conservative. This is a good example of the state of modern electrical propulsion systems:
Edit:
http://sci.esa.int/science-e/www/object/index.cfm?fobjectid=34201&fbodylongid=1539
Look at the thrust and you'll see what I mean - 68mN. Even at a light vessel - let's say 500 kg, since 1 N is the force it takes to accelerate 1 kg to 1 m/s in 1 second, you can do some kinematics assuming constant acceleration and constant mass (since propellant mass is comparatively light) to see how long it would take:
a = F / M = .068 N / 500 kg = .000136 m/s^2
from kinematics, d = r0 + v0t + 1/2at^2, where for our purposes 'd' is the distance to Alpha Centauri, r0 might as well be zero, and v0 might as well be zero (even if you include a reasonable starting velocity appropriate to a chemical rocket burn and a gravitational maneuver around Jupiter or Saturn to slingshot out of the solar plane, it practically won't matter unless we can achieve a velocity orders of magnitude higher than anything we've ever achieve before).
I won't belabor the post with algebra, but solving for t you get about 766 years to reach a distance of 4.2 light years.
If you do it again and take v0 to be 40,000 km/hr (I think this is somewhat similar to a fast probe like New Horizons), you get practically the same number - 763 years is what I got.
Incidentally, even if you increase the initial velocity by a factor of 10, and assume we can hit something like 500,000 km/hr, it's a negligible difference - still like 730 years.
2nd EDIT: In fact, from the Total Impulse (2E6 Ns) and propellant mass (82 kg) listed for that design, I was able to calculate an effective mass flow rate of .24 kg/s.... which means that for a trip time of 2E10 seconds, you're looking at a propellant mass
much much larger than 500 kg... which means all of the above calculations are
exceedingly generous...