SRM and EBC are identical except for the scale factor. Thus SRM = 12.7*A(430) and EBC = 25*A(430) as it says in the Wikipedia article. Thus EBC = 1.968*SRM for any color depth.
The SRM was designed (scale factor and wavelength which is 430 nm) to match Lovibond well in the range of interest to the ASBC which was light colored beers (this has been said already).
Lovibond tintometers are still made but they are now spectrophotometers - not visual comparators as they used to be. They take a set of spectral readings and gin up a single number from them. If we could find out what their algorithm is we could do a translation between Lovibond and SRM/EBC because those are based on average beer spectra and we know what the shape of an average beer spectrum is.
Given the situation I do not understand why anyone continues to specify malt color in terms of Lovibond. Grandfathering I suppose. Given that a modern laboratory is going to take multiple spectral measurements anyway I don't see why they can't give us 12.7 or 50 times A(430) as well as the Lovibond measurment.
In the interim if I absolutely must convert I have a fit to the EBC vs Lovibond data on Weyermann's site.
As the Wikipedia article says one cannot interconvert between the old and new EBC scales unless one makes an assumption about the spectral shape. The average spectrum is the obvious one to use but whose average? Stone and Miller's? AJ's? No one should be using the old scale.