The issue with monad tutorials is that they start at the top - trying to impart a generalizable understanding of the concept (which, short of category-theoretic explanations, requires hopelessly leaky analogies) instead of focusing on the purpose and usage of specific monads.[1][2] Abstract concepts must be first made concrete in order to understand them; in the case of monads, it's best to just look at the type signature for specific monads' "bind" (>>=) functions, as well as examples of usage, while actually using them in (permissively typed) code, rather than trying to connect burrito analogies to real life.

[1]:http://dev.stephendiehl.com/hask/#eightfold-path-to-monad-sa...

[2]:Regular expression tutorials, by contrast, virtually never attempt to explain regular languages or automata theory, which is why nobody complains about having to learn formal language theory in order to use `sed`.

A surprisingly digestible (and concrete) explaination of monads (Ch 2): http://www.cse.chalmers.se/~rjmh/Papers/arrows.pdf