Tom Goldstein has added an update to his post that responds to my points.  I’ll stand by my initial points and limit myself to one reply:

Goldstein asserts that Kagan, if nominated to the Court, “would immediately stop taking on conflicts.” Specifically, in his view, the Administration would “instruct[] [her] to recuse herself from ongoing participation in litigation during the pendency of her nomination, recognizing that for the Court’s sake it doesn’t make sense for her to generate further recusal obligations.” 

This is rich.  Kagan, in Goldstein’s view, would be recusing herself from a chunk of her current work as SG in order to enable her to participate more fully in future cases as a justice.  That’s a novel view of recusal, which properly turns on existing conflicts, not on anticipated future ones.  It also would reflect an unusual degree of presumption of confirmation on the part of a nominee.  And it undercuts Goldstein’s broader argument that recusals wouldn’t be large in number:  If they wouldn’t be large in number, why would the Administration tell Kagan to change the way that she’s doing her job in order to avoid triggering more recusals?

Yes, I realize that Chief Justice Roberts and Judge Alito didn’t take on new cases once they were nominated, a result, as I understand it, of decisions on the part of the chief judges of their courts, who understandably wanted to accommodate the demands that the confirmation process places on the nominee and who would want to ensure as much as possible that decisions were made by full three-judge panels.  There’s zero reason to believe that helping the nominee avoid future recusals played any role in the decisions.  (Oddly, Goldstein asserts that the Obama administration “ultimately settled on [Judge Sotomayor] immediately stopping participating in matters.”  That would have been a matter for the chief judge of the Second Circuit—not the Obama administration—to determine, though I’m sure that the chief judge was happy to accommodate Sotomayor’s request.)