There actually is some precedent, albeit persuasive precedent only, for the proposition that redistricting should only occur after the decennial census. First, some factual background:
In 2002, after the 2000 census, Colorado's legislature failed to perform its duty under the state constitution to redistrict after Colorado had gained a House seat. A suit was filed that sought the courts to redistrict in light of the legislature's failure to do so; the courts did so, and the 2002 elections went forward with the districts as drawn by the courts. In 2003, the legislature finally got off their ass and drew new districts. Another suit sought to enjoin enforcement of the new legislatively drawn districts and instead keep the court-drawn ones in place. The Supreme Court of Colorado struck down the new districts and kept the court-drawn ones, intepreting Article V, Section 44 of the Colorado Constitution (which says that the legislature is supposed to redraw the districts after each census) to mean that the ONLY time redistricting could happen was after the census. Therefore, the Court found, the districts drawn in 2003 were invalid. See Salazar v. Davidson, 79 P.3d 1221 (2003). There is now a challenge to that finding pending in federal court, based on U.S. Constitution, art 1, Section 4, which gives state legislatures exclusive power over Congressional elections; the argument is that art 5 section 44 of the Colorado constitution, as interpreted, is unconstitutional under the U.S. Constitution because it takes control of Congressional elections away from the Colorado legislature. See Lance v. Dennis, 546 U.S. ____ (decided February 21, 2006) (holding that the current suit is not procedurally barred by the prior state court case). New Hampshire has also found that Congressional redistricting can only happen after the census pursuant to its constitution, although under a similar scenario as described above that also occurred there, the Supreme Court of New Hampshire found that their legislature didn't lose its chance to do its once-every-10-years redistricting just because the session expired, and therefore allowed redistricting done at the 2004 session. See In re Below 151 N.H. 135 (2004).
So back to Georgia. All of the above is to say this: Georgia's constitution has a similar provision with regard to state legislature redistricting. Georgia Constitution, Article III, Section II, Paragraph II (The apportionment of the Senate and of the House of Representatives shall be changed by the General Assembly as necessary after each United States decennial census.) So, if the Supreme Court of Georgia interpreted that provision the same way that the supreme courts of Colorado and New Hampshire intrepreted their provisions regarding Congressional elections, they would find that the ONLY time redistricting could be done is as necessary after the census. What's more, if the Court did rule that way, the federal case in Colorado would have no effect, because the U.S. Constitutional provision regarding Congressional elections would not come into play when dealing strictly with Georgia legislature elections. Such a finding would invalidate the recent rending of Athens in twain.
Now, what are the chances that the Supreme Court of Georgia will rule that way? Your guess is as good as mine. It would be a case of first impression as far as I know in Georgia, and the Court has some new personnel that I don't really have a good read on yet.