>>3982904This is another example of DxOMark being misleading. Low-light ISO score shows a nearly 50% difference but they are damn near identical to my eye in DPReview's test scene. (Note: you have to choose the 5300 in DPReview's tool which according to DxO is even a bit better than the 5200.) Either of these cameras will work well to ISO 3200, or 6400 in a pinch, assuming proper exposure. Neither is usable beyond that.
The portrait score is irrelevant because you will never see the difference.
The landscape score is the ONE place where DxO gives an accurate impression. The 7D is just shy of a 12 stop DR, while the D5200 is practically 14 stops. What does this mean in the real world? When shooting bright sky/shadowed foreground landscapes you will need to be more careful with exposure on the 7D, and you'll likely want to apply a little NR to the pushed shadows in post. Sun in frame will almost certainly mean blending two exposures, on either camera, unless the sun is attenuated (i.e. clouds, fog).
You would never choose between these cameras based on DxO. First question is: what system of lenses do you want to be in? If you want Canon EF, the D5200 isn't going to help you. If you want Nikon F, the 7D is the wrong choice. Keep in mind with the D5200 you really only want F lenses with built in AF motors. The 7D will mount/use any EF lens.
If you can't decide between mount/lens system and are stuck on the bodies, then the 7D is an 8fps sports/wildlife beast, built like a rock and weather sealed, with great controls and customization. The D5200 is a consumer body with consumer level build/controls/customization. But it does have a few more MP and a little bit better video.
If you will be doing any form of serious flash photography, the 7D mops the floor with the D5200. 7D can high speed sync and can remote control Canon flashes from the rear screen using the built in popup. The D5200...cannot.
