Mathematically speaking, these grade boundaries are more or less accurate. According to this set of grade boundaries, only 3% of all candidates who took this course got 191 or above. If this course was taken by all 700,000 or so GCSE candidates, then that would mean around 19,000 people would get a grade 9. Considering that around 35,000 (5%) students normally get A* from the old specification, and around about 15% of all GCSE candidates get a grade A or an A* previously, which is about 105,000 students. Now, 20% of people who get at least a grade 7 will get a grade 9, which is about 21,000 students. So it is fairly similar.
However, this course was only taken by a much smaller number of candidates - mainly those in private schools. Using basic maths, it's clear that these boundaries are going to be less accurate than if this course was taken by all 700,000 or so candidates. In other words, the grade 9 boundary will probably be higher than this, most likely. In addition, this CIE IGCSE course and the other 9-1 courses most likely had varying levels of difficulties, thus why it isn't safe to assume that these marks can be used to accurately predict what the boundaries will be.