There is a relationship called image scale, that depends on the pixel size of the camera chip versus the focal length of the telescope. In an extreme case, if you don't match the pixel size to focal length "properly" you could end up with only part of your target on the chip, or the whole target will fit but fill so few pixels that you don't get as much detail in the image as if you'd chosen a better image scale.
This means that you should match the size of the targets you most often image to a "proper" image scale and then be sure you have a scope/camera combination that supports that scale.
However, with the very small chips we're talking about one extreme of the scale, and with the very large chips with tiny pixels we're talking about the opposite end of the scale.
My interpretation of your original question was whether the small-chip cameras have applications at all, rather than applications for which they're better than larger-chip cameras.
Just to say, for example:
Kate:
... the big one can do everything just as good as the small ones but most things they can better ...
is not sufficient. If you want to put a large-chip camera to its best use for a given target diameter, you want to spread the detail in the target across the largest number of pixels, with the best quantum efficiency, at the best image scale, for that particular size of target. While you can make some adjustments (such as binning), there are only a few choices that are appropriate for a given scope/camera combination that result in the best target 'coverage' ... where "best" is defined as the desired detail for a given range of exposures.