Recap: A point of inflection of a nice function y is simply a minimum or maximum of y'.
You already know how to find mins and maxs of nice functions, so you just apply the same knowledge to the function y'. That is, you look for points x such that y''(x) = 0 (at this stage x is a
potential point of inflection of y), and then you test to see of x is a min or max of y' by (1) applying the second derivative test to y' at the point x, or (2) by substituting x-values either side of your potential point of inflection into y' to get a feel for what the function is doing near your potential point of inflection.
Method 2 is inherently dodgy since it doesn't prove anything on its own, but will probably be good enough for any function you're likely to see and any examiner you're likely to put a test in front of. Except in special cases, method 1 has the potential to actually prove that your potential point is a point of inflection. The special cases for which this method doesn't work, and certain proof one way or the other on whether your potential point of inflection is a point of inflection, are easily taken care of by a simple generalisation of the second derivative test, but this generalisation is not considered 'HSC knowledge' (though it is a simple consequence of it).
In the special case that your point of inflection x satisfies y'(x) = 0 it's called a
horizontal point of inflection, simply because the tangent line to the function y at the point x is horizontal. If you would like a more 'geometric' feeling for a horizontal point of inflection, it is not just a point where the function 'slows down' and 'speeds back up' or vice versa (a point of inflection), it's a point where the function actually slows down to a complete stop before continuing to increase or decrease as it was before. If you want pictures to help with that, the functions f(x) = x^3 and g(x) = sin(x) at the point x=0 should help you out.
Why is a HSC 2011'er thinking they know more than an engineering student?
I don't. I just knew that this engineer was wrong in this particular instance.