I find it much easier to see what is going on when selecting λ-calculus instead of Δ-Nets. E.g. for the mandatory Y Combinator,

λf.(λx.f (x x)) (λx.f (x x))

for which the difference with

λf.(λx.f (x x)) (λx.f (x x) f)

is very clear, whereas with Δ-nets the difference is more subtle. I guess it is because the visualization has more information than with the λ-calculus.