In my opinion the both aspects are important. The comparison of the morphology of individual developmental stages is able to indicate specific damage caused by the treatment; the area covered by the cells represents the size of this cell type and its count. The size could reflect the cell condition and function, the count the time period of this stage presence in the tubuli (e.g. the spermatocyte I exists for much longer time period than the spermatocyte II, therefore it will be found much more numerous).
Pavel is right, both could be important. However, sometimes the damage is so intense that the stages and cell type are dificult to differentiate between tubules.
What happens with those tubules with unrecognisable cell types, and therefore unrecognisable stage? Where do you put them?
If you can not recognise stages for the intense damage, you could also focus in between tubules; Leydig number, separation of tubules, edema, dettached cells, tubule diameter and epithelium height, etc.
With what type of insult or injury are you working?