If you take the whole family of such experiments into account, then there are two main results/uses. These become clear in ghost imaging. Here you have the same setup, but use one large detector and one small detector which can be moved. If you now place some object in the beam pointing at the large detector and scan the small detector, you will find an image of the object in the coincidence counts although the detector behing the object is not position-sensitive at all.
The two main results of this experiment are as follows:
1) The image of the object provides superresolution. That means the resolution is not limited by the standard diffraction limit.
2) This superresolution is a result of the perfect anticorrelation of the entangled particles. Therefore the appearance of such superresolution can be interpreted as a criterion to identify entanglement which is somewhat easier to realize than Bell tests are.