Disclaimer
I was unaware of 2-9-3, but performed a simple internet search which brought up a an open-access paper from 2001 refuting the claim, and nothing more recent, suggesting that the refutation was valid. I mention this both to make it clear that although, like the poster, I am not an expert in the topic one can get considerable mileage through standard search techniques. (I do acknowledge that it does help if you know you can get access to papers through a University Library, open-access or not.)
Refutation
The claim was actually made nearly two decades ago in a paper by Vreeland et al. in Nature 407, 897-900. The date is relevant because at that time full-length sequences of bacterial genomes were only just appearing.
The paper refuting the claim is entitled “The Permium Bacterium that Isn’t” by D.Graur and T.Pupko and was published in the journal, Molecular Biology and Evolution 18 1143–46 (2001).
The authors put forward various indirect arguments for doubting the claim but they based their refutation on sequence comparison between so-called Bacillus permeans (2-9-3) and other bacteria. The first argument was that the 99% similarity of B. permeans to a contemporary halophilic bacterium, previously known as Bacillus marismortui, subsequently renamed Salibacillus marismortui, implied a relatively recent divergence from the latter. To rebut the argument that S. marismortui had also been trapped in salt for millions of years until recently a comprehensive comparison with other bacteria was conducted resulting in the phylogenetic tree shown below which was incompatible with the bacterium having not evolved in the last 250 MYr:
An independent analysis by Nickle et al. in Journal of Molecular Evolution 54, 134 (2002) came to a similar conclusion.
I am not expert enough to judge the methods of determining rates of evolution and relative distances between species, but it would appear that they are accepted by those in the field.
