The data-gathering and visualization capabilities of ConnectorDB are pretty much ready for release. All that is left is fairly extensive examples and documentation, and 0.3.0 will be released. I have been testing the newest version for the past couple months, and have not found any serious bugs yet.
I have actually found that the data itself is not enough to gain the insight I was hoping for, so my focus has shifted to the ML portion of the project - a completely separate AI which performs analysis of the data.
This means that the visualization/data gathering are now secondary concerns - they work well enough for now to bootstrap the ML algorithms, and with 2 years of personal data, I hope the uniqueness of my training set combined with some clever predictive capabilities will lead to interesting results.