I'm not a betting man but I will bet it fizzles and dies because, if the article is anything to go by, someone really doesn't understand the problem. The last few paragraphs of that article are IMO laughable, also quite insulting to schools!
The biggest reason the schools census 'works' is because school funding is predicated on the submission. That's the key motivation for the schools to do the work and the key reasons the data quality can be relied on in terms of the broad brush stuff - pupil numbers. What motivation is there for them to feed everyone else quality 'master' data? Is there even such a thing as 'master' data in some kind of absolute sense? It might seem an odd concept, but there might be more than one truth out there. So for instance, A child's surname might be different for a given context even in the same moment, they are known to one school as X but known to social services as Y. Or schools might just record something about a child almost as a place holder - "Will bring address tomorrow" might appear in the school systems address - should that overwrite the DFE information, which having beeg sourced from the feeder primary is at that moment, 'better quality'? Many of the problems are not easily solved with automation and having been generally given more autonomy over the last decade, why do you think schools will generally put effort into arbitrary data quality standards where it is not absolutely central to their operations?