In the general setting A-star is optimal. JPS (and JPS+) can speed it up if you have additional information about the topology of the search space: 1) search is on a rectangular grid, and 2) movement between adjacent grid squares has uniform cost. In that case, you can take advantage of certain properties/symmetries of a rectangular grid to safely "skip" parts of the expansion that A-star would normally do, while still being guaranteed to find the optimal solution.

The JPS paper from a few years ago explains in more detail: http://users.cecs.anu.edu.au/~dharabor/data/papers/harabor-g...