In a variety of ways, it would have been an extraordinary intrusion into state constitutional issues for the US Supreme Court to have blocked the Pennsylvania Supreme Court's order in this matter. In fact, it's somewhat alarming that anyone would have even sought such a stay from the federal court.
The PA Court found that the redistricting map that the state legislature drew in 2011 defies the traditional criteria for district drawing, namely the criteria of compactness, contiguity and respect for the integrity of political subdivisions, and thereby violates the Free and Equal Elections Clause of Article I Section 5 of the state Constitution. Expert witnesses for plaintiffs showed, and the Court agreed, that many of the districts have unusually low compactness scores; several of the districts are contiguous only by virtue of a single building, and various municipalities and communities were split, sometimes into multiple different districts, for no discernible reason other than in an effort of partisan gerrymandering.
This case is entirely unlike the challenged Wisconsin redistricting plan that is the subject of
Gill v. Whitford, which the US Supreme Court will review later this term. Most notably, the Wisconsin districts apparently do not defy traditional districting criteria, and the plan is being challenged primarily on a proportionality measurement, termed the Efficiency Gap (EG), where the proportion of Democrat vs. Republican voters did not elect a similar proportion of Democrat vs. Republican representatives. From the beginning of partisan gerrymandering challenges, the Supreme Court has repeatedly held that such (dis)proportion does not amount to unconstitutional gerrymandering. Additionally, there are multiple questions as to whether the EG is even an appropriate method to measure such disproportion, and other such issues. For instance, the EG is a statewide (not district-specific) measurement, thus cannot show that any particular district was gerrymandered. Election results can have an EG far from zero (proportion) even for district maps that were drawn with overtly non-partisan intent. Districting plans can flip signs (from positive to negative) from election to election during the lifetime of the plan. And the EG treats votes for an unopposed candidate as “wasted” votes, thus suggesting favoritism toward the other party, which is illogical.
I briefly discussed the issue of partisan gerrymandering in an OP last year:
How Can One Determine an Instance of Partisan Gerrymandering? Regardless of how common and constitutionally offensive partisan gerrymandering is, it isn't an easy issue for a court to discern. Indeed, this is exactly why to date the Supreme Court has found that it is unable to articulate a judicially manageable standard for determining when district map-drawing is a case of unconstitutional partisan gerrymandering.