Is there anything significant that could be learned from DNA and applied to traditional computer storage methods? Or is this efficient something intrinsic to DNA that can't simply "be transferred"?
The issue, at least from the last time I heard George Church speak (~Fall 2016), is that the interface for DNA storage systems would have to be radically re-imagined, since we can't go between Digital <-> DNA anywhere near the order of magnitude as we can go DNA <-> DNA, or Environment -> DNA.
Something like genetically engineered bacteria that live on the walls and have special proteins that sense and encode the environment, and count the number of unique human genomes they encounter, like a time series database (think PipelineDB) may be possible.
[1] I'm pretty sure it's based off this paper, but I'm not 100% sure. https://www.sciencedaily.com/releases/2009/07/090715131447.h... Basically, the idea is that for the energy used to copy DNA, DNA polymerase is very good at only inserting the "correct" nucleotide.
DNA mastery is important on its own terms as biology mastery, but I wouldn't be surprised there is no point in the future in which it is the optimal mechanism for storing generic data.
The estimated half-life for DNA is apparently 521 years [1] (still quite a bit longer than 50 years).
[1] https://www.livescience.com/38150-dna-degradation-rate.html