The age range of 12 months between oldest/youngest in class is not the problem IMO. And even if it were, then extending that age range to 15 or 17 months clearly wouldn't be the solution!
Having a cut-off date at the end of August isn't the problem either. If it were, then moving the cut-off date to any other date would just be moving the problem to other children.
Having a 'school' start age of 'too young' at 4 IMO is not the problem either. In the US I understand they start later, and yet have the same discussions about the youngest being at a disadvantage. In contrast, in France kids go to the 'maternelle' from 3 but I haven't yet heard of any 'summer born' equivalent debates (which doesn't mean they don't exist, I'm not very clued up on France).
But clearly there IS a problem, and it is big enough of a problem that despite being recognised, and despite teachers/schools being held accountable for how well their summer born children are doing, the problem persists all the way through school/education.
IFS report 2007
(Statistically! My own late August born boy is doing well academically and socially in his older-than-average class. But the plural of anecdote is not data!)
So what is the problem? I think the problem is what happens to children when they start school. Whatever age they are, 3, 4, or 7 - if the 'curriculum', the format of learning, the structure of the day, and the achievement expectations match the children's developmental stage, then they will be alright.
In contrast, in our schools in England only about 2/3 of children achieve a 'good level of development' by the end of reception (this has improved in recent years from only just more than half). To me that says that the expectations for what is a 'good level of development' are too high.
By the time children go into Y1, the format of instruction (in most schools) becomes very 'school-ish' which is fine for mature 6 year olds perhaps, but not so for only-just-5yos (who may already be disadvantaged by having essentially 'missed' the whole of reception (despite attending) due to not having been ready).
KS1 SATS are meant to be a challenging test for 7 year olds (with standards having been raised from what was expected of 7 year olds before the 2014 curriculum) - but a full quarter of children sit them when they are still 6.
Later on we have things like the 11+ being sat by just-turned-10 year olds (or in some cases, by children who are still 9). Here I find it harder to argue that age alone explains the summer born disadvantage. More likely it is a cumulative affect of negative school/learning experiences.
So IMO the problem lies in the expectations, and the format of instruction, not being in line with the children's ages/stages. Or in other words, let the children learn through play throughout KS1, get rid of KS1 SATS, and adapt EYFS expectations to match the children's effective ages.
Or what this excellent opinion piece based on recent research argues, don't try to teach the children reading and writing before they have learned to talk and listen properly ...