It all depends on the question you want answered - if you are just looking at how a particular protein changes with treatment then western, ELISA, ICC/IHC, dot blotting (equivalent to western) and flow will all work fine and are simplest. Searching for unknown responses often requires some proteomics and bioinformatics (e.g. for localization signals)
The first 5 are all antibody based methods, and as such rely on good antibodies to give you the results, so can be incredibly sensitive and accurate grading right through to terrible. However, these are relatively cheap
Proteomics can be good, quantitative and reasonably sensitive (getting better all the time too), but quite expensive.
Mass-spec identification can be good too, but it isn't very sensitive (basically you need to see a band on a coomassied gel for it to work well), and is moderately expensive.