It's worth noting that this proof would not have established the maximum area at all without the equality part.
Leaving out the equality part, the proof establishes that the area of any triangle with semi-perimeter s cannot be more than s2 / root 27, but that doesn't mean that a triangle with that area is possible. It is like my saying that a circle of radius 1 fits entirely with a square of side length 4, and so the area of the circle is at most 4 square units - the proof establishes an upper bound. It is only the equality part that shows that a triangle of the stated area does exist and that it is equilateral. The proof does not formally address that any non-equilateral triangle with that perimeter will have a smaller area, though that is true from the AM-GM as equality occurs if, and only if, the terms are all equal.