it was not designed (in part) to provide resistance against the 282T mutant, that was a lucky coincidence which was not expected.
The paper you referenced was published in March 2011. Here's a VRUS PR from April 2010 which proves the company was aware that PSI-938 was active against 282T.
Pharmasset's purine analogs retain activity against the S282T mutation associated with in vitro resistance in other nucleoside/tide analogs in development, and are metabolized to the active triphosphate form through a different phosphorylation pathway than the pyrimidine analogs.
Other purine analogs under development are unable to make such a claim.