Having separate layers of abstraction is a good approach in general: see "Abstraction Barriers" from SICP (https://mitpress.mit.edu/sicp/full-text/sicp/book/nod

Having separate layers of abstraction is a good approach in general: see "Abstraction Barriers" from SICP (https://mitpress.mit.edu/sicp/full-text/sicp/book/node29.html https://mitpress.mit.edu/sicp/full-text/sicp/book/node29.htm...). Lasagna code is abstraction done wrong, with possibly too many layers or leaky abstractions (like in a real Lasagna).