In my experience there often is not a good reason to keep images in 16 bit. It depends on the image, but in most cases there is no color information in the added bit depth to justify the added storage needs. There are times we keep images 16 bit, if we can see that image information requires it. The thing is, it's not an issue of the image having more color, but more gradient color, especially in a single color range.
For example, an image that has a lot of gradient sky may benefit from remaining 16 bit. Large swaths of gradient sky, especially, can benefit from remaining 16 bit. So some photography with large swaths of subtle gradient can benefit from a higher bit depth. Some paintings that have a lot of subtle gradient across large areas may benefit from 16 bit files. But in most cases we will reduce an image to 8bit after we're finished editing it.