The black birds of southern Africa have indeed been suggested as a possible separate species, to which C. serratus would apply - but they are currently (almost?) universally regarded as a colour morph.
(This was originally a BSC split involving two partly sympatric forms, btw : invoking the PSC to "accept" it is a bit odd, I would think.)
In the currently widely recognised treatment, serratus is applied to southern African populations regardless of their colour; jacobinus to S Indian/Sri Lankan birds, and pica (type locality in Sudan) to all the rest.
The three are said to differ mainly in size and in the frequency of the colour morphs.
Nominate jacobinus are smallest; serratus are rather large; pica are large too but vary geographically in both size and structure (wing:tail ratio), with some populations averaging smaller than serratus, others larger.
Roselaar in Cramp (1985) recognised four colour types - pied, intermediate, grey, and black - but variation is obviously rather gradual and slight between the three former. Nominate jacobinus are pied; within pica, most Indian birds are pied too, while most Afrotropical breeders are intermediate (Roselaar wrote that the occasional black individual may also occur; HBW, on the other hand, suggests that black birds from S Africa may wander to W Africa - I suspect these are two interpretations of the same evidence); serratus are grey or black.
Serratus also differ in their preferred host (bulbuls or Fiscal Shrike, versus Turdoides spp. in other populations) and in egg colour (white instead of blue, but white eggs also occur rarely as far north as Mali and Kenya; see HBW).
This is a bit of a guesswork, but :
- Separating pica alone is odd indeed, because this would leave together two populations that are geographically disjunct, and about as different as two Jacobin Cuckoo populations can be. I would bet that M&S had serratus in mind.
- There is no consistent size difference between pica and serratus; differences in the frequency of colour morphs alone are often not considered a valid base for delimiting subspecies (see the monotypic Gyrfalcon); host preference and egg colour are interrelated in cuckoos - egg colour typically tends to match that of the host - and are known to vary from female to female within single populations in other species - e.g., Cuculus canorus - so their taxonomic relevance in this group could perhaps also be questioned. I would suggest that R&A do not recognise pica as distinct from serratus, in which case serratus becomes indeed the form occurring in N India.