That doesn't really follow Erebus. You are presuming that reading age should correlate exactly with SAT expectation, and it probably doesn't. Also, the difference between 8.5 and 10.1 is small enough to probably fall within the realm of "average". And these things are just not an exact enough science to be able to make solid statements and predictions based on such differences.
Level 4 for KS2 SAT is the expectation of national "average", so the school may be perfectly right. OTOH if you feel he could do better with private tutoring, then you're probably right - most kids could.
I agree with noblegiraffe, who put it very succinctly. The government-initiated culture of SATs, League Tables etc. has led to a generation of parents and commentators who think educational matters can be quantified, measured and predicted far more accurately than they can.
Progress is indeed not linear. Children DO stagnate and even sometimes appear to go backwards, for reasons that can be difficult to work out. I would go further and say that learning itself is not linear. It's a complex interaction of methodical teaching process with many variables of motivation, attitude, connections being made with other areas and so on. Much, possibly most, of what contributes to people really learning things is informal learning and the reinforcement of what is learnt in everyday life.
There's simply no way that a teacher can chart or predict the progress of 30 kids doing something as capricious and unpredictable as "learning", as if it's some kind of one-dimensional linear process.