Last night I was reading a paper discussing viability assays by Thiele and Back thanks to my wife who works at a university. It had one of the best descriptions of the advantages and disadvantages to Methylene Blue staining that I have ever read, and it is much better presented than I could ever write.
Vitality/Viability Assays - FRITHJOF THIELE and WERNER BACK published in Yeast flocculation, vitality and viability : proceedings of the 2nd International Brewers Symposium
Apart from the general overestimation of live cells, MB is criticized as it poorly differentiating between live and dead cells (due to partly or slightly colored cells). As a practical approach, all slightly/partly colored cells can be considered dead, which reduces both overestimation of viability and operator subjectivity. Even if this assumption is not scientifically correct (since slightly/partly colored cells show some reduction power for the stain), this approach supports the practicability of the method.
They also discuss several other methods such as cell cultures and measuring other cellular components such as glycogen Glycogen, trehalose, sterols and unsaturated fatty acids. One of particular interest to me, being an electrical engineer, was that the viable cell count can be estimated by measuring capacitance.
If anyone is interested in yeast viability, see if your library can get you this paper.