There is a considerable problem with this announcement.
Any given class of school children will vary in age by just under 12 months.
At the age of 5, somebody on the cusp of their 6th birthday will be ~20% more developed on average than somebody just turned 5.
At the age of 11, somebody on the cusp of their 12th birthday will be 8.3% more developed on average than somebody just turned 11.
You expect, on average, younger children to perform less well in tests than older children in the same cohort.
Even if results are averaged across a class, variations in the distribution of ages in a given class will affect this result.
This advantage narrows as children get older, making tests at later ages progressively fairer in terms of measuring ability, but it seems a profoundly silly thing to do for Primary School age children.