In the past, we've accepted more or less any algorithm into sunkit-image. This has worked fine as we've not had an abundance of potential contributors and thus haven't had to turn any away. However, as more potential contributions come in, this may increasingly become a problem.

In general, we should come up with a policy for what kind of algorithms we accept and which we don't. In general, I would recommend that we only accept algorithms with an accompanying publication and/or ones that are widely used or perceived as generally useful throughout the solar physics community.

Additionally, we should think about to what extent we own that algorithm. In the past, we've ingested all of the code into sunkit-image, but there's also a strong case for just providing a wrapper that exposes a consistent interface (e.g. #188).

We need to agree on what our policy is regarding the above points and then document them in the documentation.