In non-perturbative approaches to QFT the $\Lambda \rightarrow \infty$ limit shows that the QFT is well-defined as a continuous theory. However perturbative approaches sometimes have the different goal of computing $S$-matrix elements. In the latter we demand that the parameters of the theory (such as the mass, coupling constant...etc) take certain values at a reference energy scale. In other words, we treat the QFT as an effective field theory. From a Wilsonian renormalization group point of view, we can say that we are looking for a Lagrangian where we have integrated out all momenta above our reference energy scale when there is a cutoff.
What has been confusing me is that in particle physics books they still are interested in either the $\Lambda \rightarrow \infty$ limit or results independent of $\Lambda$ even in these perturbative scattering computations. Given what I wrote in the first paragraph this seems counter-intuitive. We already have a reference energy scale that gives us our physical parameters, so why do we care about the $\Lambda \rightarrow \infty$ limit? Why is having a cutoff-dependent result a bad thing if, to my understand, that is essentially what it means to demand that a Lagrangian parameter takes a certain value at a certain energy scale?
A related question is then: What role does the cutoff play in both perturbative QFT and Wilson's renormalization group? But perhaps this better suited for a separate post.