2D estimates of a population distributed in a flat sheet provides a reliable estimate of the population in the culture. It all depends on what you want to do. If the objects you want to count can be seen in a single field of view then many software packages can provide what you need. If you want to count individual cells and the population is large, then you might want to consider estimation methods. Another considereation is how easy it is to identify a cell. Does it stand out from the background? Is there a need to differentiate cells within the culture? Do the cells form clumps that need to identified as individual cells?
Can you tell us more about what you are doing?