The game I'm working on is very simple. It's a top down scroller. The camera never moves in any way, and all of the objects are in the same location on the Z axis. Since it's a bullet hell game, there are a lot of models on the screen at any given time. That being said, when I call RenderTarget2D, should I set mipmaps to true or false? I haven't noticed any difference at all so far either way.
I think you're confusing a couple of things here.
First mipmapping: mipmapping is a technique where if an object is small on the screen a small texture is used instead of the a bigger texture. This reduces texture bandwidth usage by just loading a texture that is big enough, instead of loading one very big texture. And it alleviates certain aliasing problems. See this wiki page.
Now if you construct a RenderTarget2D with mipmaps then the RenderTarget, which is also a Texture will have mipmaps. This means that the resulting image on the render target will have a full mipmap chain, not that the objects rendered to the render target will be mipmapped. Setting mipmaps to true on rendertargets is only useful if you are going to display the resulting texture and different sizes.
So the short answer is: no :)