Yeah. So when a photosite gets smaller, the total charge it can hold is also reduced, decreasing the dynamic range of the individual site. However, the decreased area also decreases the intensity, so this shouldn't be too much of a problem.
The other thing to keep in mind, especially with regards to CMOS sensors, is that there are a number of transistors on each photo-site, which decrease the fill factor (photosensitive area)/(total photosite area).
Moore's law has largely been successful due to the shrinking of transistor sizes. Just because you shrink a photo-site doesn't mean you actually shrunk the photosensitive area. In fact, if your transistor sizes shrunk significantly, it may be possible to actually decrease the total photosite area, but increase the photosensitive area.
Not only would the chip have a higher fill factor, but it could also have higher dynamic range per photosite.