Are web shells changing journalism?

Interesting (well, I thought it was interesting) essay on web shells in OJR. Web shells are the navigation surrounding a story, linking to related data, resource, sidebars, backgrounders, archived stories etc. Delivering a good web shell takes you straight to the keyword/autonomy/etc debate about how you identify the context of a story. I have a problem with keywords because the context of a story changes over time: simply chucking a story about Imran Khan in the "cricket" folder might have seemed OK 20 years ago, but that story could now have an entirely different relevance now. The question with web shells is: do you build them from the story (very labour intensive: do you keep going back through your archive and putting new links to old stories?) or from the subject (restrictive: 50 stories in a subject folder X won't necessarily fit when the subject "evolves" into Y). Then there's the question of context over time: should you "snapshot" the web shell around an archived story to show the context as it stood at the time; or should you always show what you reckon to be the context now, rather than then?
(From Online Journalism Review)