Though if it weren't labeled as "Sandy Island", I would think it's just an artifact of some kind. It certainly doesn't look like any real island at that distance.
Landsat[3] never showed anything there, especially at this size. Earliest depiction[4] I found in the short time via oldmapsonline.org dates back to 1881 While this otherwise well explored map[5] of 1862 does show nothing in place.
GeoGarage[6] pretty much says the same.
This only shows how slowly low priority public datasets are updated/maintained since the french maritime authorities SHOM don't have anything in their official charts of the area.
[1]http://www.ngdc.noaa.gov/mgg/shorelines/data/gshhs/version2....
[2]https://browse.digitalglobe.com/imagefinder/navigator.do?nav...
[3]http://landsatlook.usgs.gov/ sorry no direct link possible, just got to the location zoom in and search for available imagery.
[4]http://www.davidrumsey.com/luna/servlet/view/search?q=List_N...
[5]http://www.davidrumsey.com/luna/servlet/view/search?q=List_N...
[6]http://blog.geogarage.com/2012/11/south-pacific-sandy-island...
Look at the french charts featured at geogarage, the post has added a few more SHOM charts for reference which show no Island.
http://blog.geogarage.com/2012/11/south-pacific-sandy-island...
So what does it mean?
Basically one would only need to blame Australia for the lack of funding or interest to ground-truth and update their issued charts outside of their territorial waters. The easiest explanation would be a lack of communication between Australian and French nautical mapping authorities. Second to that the Australian authorities didn't feel the need to do their own mapping by means of actively acquiring of information on the area by satellite, plane or ship for civil applications.