Why does all this matter?

If you opt to use a language that uses an interpreter, the time from writing/editing the source code to running the application is reduced. Not having to wait for your source code to be compiled could save you a significant amount of time (depending on the size of your project, the compiler and the performance of your machine).

You also won't need to generate multiple output files, to run on each different architecture. You will be able to run your source code, in the same way, on all architectures and let the interpreter handle the rest.

If you opt to use a language that uses a compiler, it is generally going to be more performant. It won't have the additional step of having to translate the source code on the fly, because all of the source code would have already been translated.

One final note; just because a language uses an interpreter or compiler, doesn't mean it can't use the other. Usually, there will be a standard or accepted method to use, but it is possible for someone to write an interpreter and/or compiler for any language.