Why not?

Piracy is a problem for content generation because somebody needs to get paid to generate the content. Elsevier does not generate content, it's a (parasitic?) gatekeeper using network effects to extract rents.

If content generation would be healthier and more vibrant without Elsevier, then piracy seems an appropriate and just solution. I don't see content creators lining up to defend Elsevier; quite the opposite.

(It's possible that someone could argue that Elsevier adds significant value for their cut, but I haven't heard anyone make that claim)

I get the idea that skirting the law is not something we should generally encourage as the solution, but let's not get carried away and act like copyrights are some divine commandment and grad students using scihub are a bunch of evildoers.