The Heisenberg uncertainty principle indeed says that we can only achieve a position uncertainty of delta x >= hbar / (2 * delta p). However, this is of course a lower limit on the uncertainty of the position, and hbar is extremely tiny. In practise, we have many other classical noise sources that limit us far before we can ever reach a point where we are Heisenberg-limited.

The noise due to the Heisenberg uncertainty principle would only become "visible" when all of the classical noise sources are reduced such that the classical contribution to delta x * delta y is less than hbar / 2. I don't think this experiment reached such levels of precision