Of course, there's going to be some toxic DNA sitting in the sequences they are making, and it will be to the advantage of the bacteria to spit it out, they will find a way to do it with recombinases, even the ones with the most pernicious recombinase (recA) knocked out.
There are two problems with bacterial data storage. The first is information retrieval. Running a sequencer is no fun.
The second problem is genetic recombination, which is what they are using to 'achieve encryption'.
http://en.wikipedia.org/wiki/Genetic_recombination
Of course, there's going to be some toxic DNA sitting in the sequences they are making, and it will be to the advantage of the bacteria to spit it out, they will find a way to do it with recombinases, even the ones with the most pernicious recombinase (recA) knocked out.