I think that in a true catalogue (which Spink is, not just a price guide) proofs and patterns should feature.
However, I do think that there are currently two distinct groups of collectors - the more numismatic kind, who tend to prefer the vast scope of predecimal coins, and those who collect decimals and concentrate more on the present realm of currency coins. So the decision to hive decimals off into a separate book makes logical sense. Our hope should be that those who start with decimals may move on to predecimals. A useful precedent is coin collecting pre-war: that was predominantly split into ancient, hammered, and early milled; anything post-1816 was seen as somehow 'modern' and inferior, and it took the complete disappearance of silver to change that.
One thing I believe could usefully be done though : just as all farthings from 1918 to 1925 are grouped into a single entry, they could do the same with all milled coins: give a base set of prices for that type and list all the dates it includes; then list significant variants and varieties as separate entries below that. For example, with George VI brass 3ds - give a base price and include 1937, 1941/2/3/4, and 1952 within that, then separate listings for all other dates. That would save some significant space and also resolve the very trivial price differences we see in modern ranges, which actually bear no relation to rarity and are merely a continuing and now inaccurate throwback to the late 60s, when the comparative scarcity of even modern and common coins was considered worth noting. Since the Great Meltdown, that has never been properly revised.