The big problem with finding the sign of the cosmological constant is that regardless of whether it is positive, negative, or zero, its magnitude seems to be very small. This means that while we have done a great job of pushing down our error bounds, they still contain both positive, negative, and zero values for the cosmological constant. As an aside, note that if the cosmological constant is truly zero, then no matter how carefully we measure it we will always have error bounds that contain both positive and negative values, and so we would have to find some other way to conclude that its value is zero.
A comment to the question above asks why the fact that the universe's expansion is accelerating doesn't prove that the cosmological constant has positive sign. The reason for this is simple: there are different reasons why the universe's expansion could be accelerating. If the accelerating expansion is due to a cosmological constant, then such a constant would have to be positive. This is the foundation of the $\Lambda CDM$ model of cosmology, but it isn't the only option. Dark energy does not have to take the form of a cosmological constant; it could be something else entirely. If this is the case, the fact that the expansion of the universe is accelerating has no bearing on the sign of the cosmological constant.
Because we know so little about the nature of dark energy, we can't make any clear-cut conclusions about whether or not it is simply a cosmological constant.