Mathematics papers (dissertations included) are meant for active researchers. Perhaps that implication really is obvious to an active researcher in the field. The fact that it is not obvious to someone outside of its intended audience is not a mark against the paper. If I said to someone in machine learning that by assuming zero-mean Gaussian priors for the coefficients, we are encouraging small values of the coefficients, that would be really really obvious to someone in ML/stats/data science (I've restated a really basic statistical learning concept), but not to a random educated person.