Did a bit more digging earlier today and found a few interesting things to consider although the death certificate and that birth certificate needed to confirm the correct family so you don't go off on a wild goose chase after the wrong family or even the wrong branch of the right family.
I think it's possible that granny wasn't called 'Cath' but her great-grandmother could have been Catherine.
Searching for Mullark as the surname doesn't bring up many entries in civil index (births, marriages, deaths) because, like many Irish surnames, different spellings were used at various times. So, searches have to include Mulark, Melark, Malark, Mylark, etc.
Now, if Margaret was born c1925 to an unmarried mother named Mullark (or a variant of Mullark) then she should appear in the 1911 census (aged at least 14 years) and possibly in 1901. Unfortunately, some of the Rooskey D.E.D. seems to be missing from online 1901 census (whether just one that missed being microfilmed I'm not sure).
What I have noticed in records I've checked so far is a strange lack of suitable girls the right age to be Margaret's mother (unless she was away from home in 1911 and her surname hasn't come up in my searches). So, I have a particular suspect for who I'm constructed a nice family tree but will wait until confirmation of the name of Margaret's mother. If my guess is mistaken then I have little mini trees of the onnected families in the area.