The route was invented in Russia - desomorphine was used in Switzerland in the 1950s as a fast acting, short duration, potent analgesic. The main problems with the Russian route are that the yield is so poor, they cannot isolate the active AND codeine undergoes rearrangement at the allylic hydroxyl producing pseudomorphine, iodinated side-products and god alone knows what else. I have a Russian paper on it somewhere and yield is 2-2.5% hence 100s of mg of codeine are required to produce a hit (5-10mg).
I think one solution (a double-edges sword) would be to use dihydrocodeine. Reduction of a plain, secondary hydroxyl is much higher yielding AND no rearrangement can occur. Of course, that would mean a lot more overdoses. DHC can also be demethylated much more easily. AFAIK, the best use for codeine is it's rearrangement to hydrocodone (there is a 2013-4 patent giving 95%+ yields). Hydrocodone can be demethylated as easily as DHC.