Underfire, that diagram has to be wrong; every PAPI on an ILS runway that I know about has the same angle eg 3°. That diagram clearly shows the PAPI path is much less than the ILS GS.
The diagram is coincident, both the ILS and PAPI are set at 3°. The diagram shows where the points meet, when and how they diverge.
The optimum is 3°, but it is often higher/lower depending on obstacles and intended use. Frequently, there is a difference in the GPA, because of the obstacle clearance surfaces and the respective procedure.
In the US, I actually find the non-coincident comment very often, the recent events at SFO have brought this to the forefront.