I haven't seen the tablet tests over here in the states, but I think that the tablets have the potential to be more accurate.
With the liquid tests there are two user-induced errors: Putting an incorrect amount of water to be tested in the tube and putting an incorrect amount of reagent(s) in the tube. The first happens because it isn't always easy to get the water line to match the line on the tube. The second happens because not every drop is exactly the same size -- a lit depends on how hard you squeeze the dropper.
The tablets eliminate one of those errors -- the amount of reagent error. One tablet should be the same size each and every time. So you know you get the same amount of reagent each and every time.
I terms of accuracy, this comes down to how easy it is to determine the color of the reagent in the tube with the standard on the card. Ideally, you'd like one that has a very sharp color change as you go from low to high concentrations. This doesn't always happen -- API's nitrate test is almost worthless when the nitrate starts getting above 40 or 50 ppm.
So, if someone could post a pic of the standard cards for the tablet test, the range of colors should show how accurate the tests can be. And, if there is a good color change over the range, I think that I'd prefer the tablets because it eliminates one of the two common errors when performing the test.