planet - Where might a semi proficient amateur analyst participate in meaningful astronomical efforts

I haven't managed to formulate a coherent single answer, but here are several suggestions for where to find some hints of current topical research where you might be able to contribute.

Open source projects

Astronomy is increasingly using large open source projects, many written in Python, which itself free. For example, the Astropy project is trying to create an extensive Python library for astronomical data manipulation, so you could try to contribute some of the desired features in a project like that. (There are other, more substantial, projects, especially in various kinds of modelling, but they require an understanding of the physics and are often big chunks of code, written in Fortran.)

Review articles

If possible, I'd suggest looking at review articles (e.g. in Annual Reviews in Astronomy & Astrophysics). Though many of the articles are probably behind paywalls, most of the recent ones should also be available on the arXiv. Similarly, you can also try searching arXiv for lecture notes from summer/winter schools.

Departmental PhD project listings

In a similar vein to contacting people, you might find that potential projects are listed online, in which case you'll see what kinds of things people would like to do. For example, so quick Googling netted me information at Manchester, St Andrews, QMUL, and UCL. (Clearly Google thinks I'm in the UK...) While it probably doesn't make sense to actually try to carry out these projects, they might give you a better idea of the sorts of things that need doing.

Observation projects with public data

I'd particularly watch out for anything that involves data-mining, since that mostly involves spending time crunching some of the huge datasets available. I'm mostly aware of projects in the time domain (e.g. OGLE and WASP) but there are also larger projects like SDSS that I think have more data than people to sift (intelligently!) through it all.

I'd note here the special cases of Kepler and it's continuation K2. In these cases the actual analysis of the cameras' pixel data is still an open question, especially for K2. Any clever progress on automatically reducing the data better would be a boon in that field, although several active research groups are also working on it full time.