One of the things that takes practice and observation is getting the blend correct - not in terms of masking, etc., that is a mechanical process. The fact that you are making a blend is because, typically, there are generally two different levels of scene illumination that challenge or exceed the camera'/raw converter when trying to capture and render that scene from a single shot. So, you take two exposures and blend them, getting the best pixels from each shot and making a composite image from the two. Or you process the same image twice, etc.
Great. So far, so good.
However, in doing so you are compressing the tonal range of the scene to get all of that dynamic range into a single image. Moreover, you are, more or less, compressing the tonal range arbitrarily. This leads to unnatural images. Think about the relative levels of scene luminance and how you can reestablish that once you blend the two sets of pixels into a single good working set. Make notes (mental or otherwise) when you shoot the scene, take meter readings, etc. to get a feel for how relatively bright one critical area is compared to others. This will help you achieve a more natural result when you reestablish contrast.
If the scene were uniformly bright, then exposure blending would not be necessary.