In detail: Beck's original, above, and Noad's twist on it, below…
Nothing is exactly sacrosanct, but some things just don't need redesigning. Coca-Cola just doesn't need a new logo, however much one might see a design challenge. It applies across the creative scale: the Mona Lisa doesn't need repainting, 2001: a space odyssey may be 40 years old but it doesn't need a remake. Occasionally the unexpected happens – in my own view, Gus van Sant's shot-for-shot remake of Psycho (but in colour and set today) hit some very interesting spots about film-making, and was always intended as a comment or an exercise, not an improvement.
In the design world, Harry Beck's original London Underground map (although adapted for the expansion of lines) does not need a redesign, however much of a valid student-style exercise there might be in doing that. (When I was at school, the art teacher set one task of copying a great, favourite work of art, in an exercise not of improvement but of realisation just how hard it was to achieve the original.)
So designer Mark Noad's effort at redesigning and hence (I assume) "improving" by making more accuarte the links between layout and geography, is, for me, most likely to stay on the drawing board.
But now, according to fastcodesign.com, it has created a small furore over what "design", a "map" or a "diagram" is. Renown designer Erik Spiekerman (who has designed the Berlin subway map) has both pointed out the London "map" is a diagram not a map and to say in a comment (in the above link), "The worst [maps] are those in this [Noad's] example, because it is neither a diagrammatic representation of an invisible system underground, nor a true image of London’s complex topography and its transport connections."
He's right about the Beck map being a diagram, not a map, but perhaps a little too literal: one definition of a map includes "a diagrammatic representation of an area…" and most users would use the London Underground map as a general map (a diagrammatical explanation with direction illustrating how to get from A to B). The accuracy to topography is essential the more worthwhile and useful a map would claim to be, of course, but centuries-old maps based on limited knowledge are still maps, even if inaccurate. And in the hands of users, simple, illustrated tourist maps are still "maps" even without exact topographical accuracy. Semantically, Spiekerman is correct – Beck's work is a diagram showing connections unrelated to exact geography – but practically it doesn't matter (to the user) as the fastcodesign.con poster above, John Pavlus, said.
However, I see Spiekerman as fully correct in his statement above about this new map not being a good example of improving on Beck's original.
Meanwhile, visitors to London might rely only on a subway diagram, but most would have a street map to compare and realise distances and relation. And Londoners also have that other bible of getting around its above-ground streets, the London A-Z map book. I personally can't see Noad's work being included in that as the tube map in the near future…