Autoclaving components of the medium separately has been done since the 1970s. Here's a paper published by my dad in 1975 in the Proceedings of the National Academy of Sciences. Look under Materials and Methods, Media. It's mentioned like it's no big deal. They had tried a number of combinations, this is what worked best for them.
EDIT: My dad has Parkinsons now. It's great to see him light up and talk about his work.
That is an awesome find. It has to feel pretty amazing to figure out that something that has been "the standard" for so long is broken in a fairly simple way.
Yeah. Like, how embarrassing.
I read somewhere else a theory of how buildup of hydrogen peroxide in the body is what kills victims of septic shock.
The only published reference I can find to that idea is http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4038812/ which was published earlier this year. It doesn't seem to have evidence based on actually measuring H2O2 buildup, only an observation that various aspects of sepsis, when seen elsewhere, are correlated with H2O2.
Amazing to know how little we knew and experimented with such a popular process used for so long. It makes me wonder what's undiscovered about other processes we take for granted.
I am no expert on this topic, but this story has a strong "bad science journalism" smell. The pattern is
(1) longstanding complex open problem
(2) stupidly simple "solution" claimed as a panacea.
Generally with this pattern, the simple solution turns out to be well known among the top experts, and really isn't a solution to the whole problem, just a tiny part of it.