This is a brief review of the relevant research evidence which overwhelmingly supports a later start to formal education. This evidence relates to the contribution of playful experiences to children’s development as learners, and the consequences of starting formal learning at the age of four to five years of age

There are several strands of evidence which all point towards the importance of play in young children’s development, and the value of an extended period of playful learning before the start of formal schooling. These arise from anthropological, psychological, neuroscientific and educational studies.  Anthropological studies of children’s play in extant hunter-gatherer societies, and evolutionary psychology studies of play in the young of other mammalian species, have identified play as an adaptation which evolved in early human social groups. It enabled humans to become powerful learners and problem-solvers. Neuroscientific studies have shown that playful activity leads to synaptic growth, particularly in the frontal cortex, the part of the brain responsible for all the uniquely human higher mental functions.

In my own area of experimental and developmental psychology, studies have also consistently demonstrated the superior learning and motivation arising from playful, as opposed to instructional, approaches to learning in children. Pretence play supports children’s early development of symbolic representational skills, including those of literacy, more powerfully than direct instruction. Physical, constructional and social play supports children in developing their skills of intellectual and emotional ‘self-regulation’, skills which have been shown to be crucial in early learning and development. Perhaps most worrying, a number of studies have documented the loss of play opportunities for children over the second half of the 20th century and demonstrated a clear link with increased indicators of stress and mental health problems.

Does the evidence support a "later start" for formal education, or does it actually raise the question whether "formal instruction" as currently practiced is ever the optimal strategy?  

Inconclusive but as Dawn points out below, late starters have significant advantages.

Is the Montessori Method considered "playful learning"?

Yes!

I'm not an elementary educator, but what came to mind immediately is Malcolm Gladwell's study of the birthday cutoff dates for hockey players...basically, the players that have the 11 mo advantage rock the game. I'm wondering now how much of these studies isolate the play (and they might, I haven't read the studies themselves--they were referenced but not cited in the article) as the element that leads to success vs. perhaps the maturity element. Studies have shown that maturity does affect school success in terms of behavior, ability to stay on task and focus. 

That was in Gladwell's book Outliers -- did he ever write a follow-up?