This new convention that the Roberts Court is developing that Presidents have to have a "good enough" reason for reversing their predecessors' executive orders further erodes the distinction between EOs and laws and is thus very dangerous.
This doesn't concern an executive order. Just as
DHS v. Regents didn't, this didn't either. New executive orders that are just that can overturn previous executive orders that were just that & - barring an
extreme precedential overturn that's not happening under a Court that Roberts/Kavanaugh control, let alone really any Court, for that matter - that'll always be the case. However, what this case legally concerned, like
DHS v. Regents before it did, was an agency action (both cases concerned DHS memoranda), & as much as the outcome of a particular case concerning agency action can suck (i.e., this refusal on SCOTUS' part to issue a stay) or not suck (i.e., SCOTUS' ruling that the attempted rescission of DACA didn't comply with the APA) depending on a particular policy in question that an administration sought to issue an agency action without complying with the APA in regards to, agency actions need to comply with the APA if the law is to
mean anything.