I have a new account. I can't comment, I can't use chat room.
This is a common complaint, but is not something that is considered broken or in need of being fixed. The small rep caps are deliberate, and provide an effective barrier against spam and misuse of additional site features.
The core site features are the ability to ask and answer questions. These are the purpose of Stack Exchange sites, and you have access to those features immediately on creating an account.
All I can do is post bad questions that are bothersome.
Don't do that. But do ask on-topic questions if you want to. Asking a question here is harder than it seems at first, because the site is looking for high quality questions that explain a specific problem well.
I occasionally would like to be able to comment on threads I see that I CAN actually contribute to, but I don't have enough rep.
Comments under questions and answers are not really like forums. They are supposed to be about refining details of a question or answer. Most often they are used to ask someone writing a question to add more details, or clarify some part of the question. On answers they might be a suggested correction or extension.
You should bear in mind that comments can get deleted, purely because they are "just comments". Some Stack Exchange sites do this routinely, with the admins cleaning up old comments as a matter of course. Data Science does not do this, but might change to doing it at any time.
Or hang out in chat and post an occasional message if I see something I've had issues with before/know how to fix.
That should be fine. Chat is more open-ended and can be social. Chat rooms spring up for various reasons - most commonly if comments start getting extended. On busy sites, chat often has constant ongoing discussion. On Data Science, we are not so busy in 2017 that you will find lots of chat.
Any ideas how to get past this without posting bogus questions? Or how to get at least 20 rep for chat?
Posting bogus questions is likely to lose you rep and will be counter-productive.
The quickest way to get 20 rep is to write high quality content in a question or answer. That can be relatively short, and you could pick an easy-to-answer question. Look carefully at the style of answers that get up-voted. They are not chatty, they don't extend questions with "that's interesting, I also found that...", they don't just link to possible answers off-site - these things are the norm on most social media and forums, but not here.
I have 6 months to live due to a degenerative disease in my spine and I've been trying to learn random things to take my mind off of it
That's a tough break. I hope you have a good support network around you. Stack Exchange sites, for better or worse, cannot really take this sort of thing into account. In some ways this is internet Nirvana - they don't care what age, sex, religion, you are, or what you are like physically. In other ways it is impersonal and unforgiving, because Stack Exchange does care about the quality of your writing and choice of subject matter, and you will be judged for it constantly in questions and answers.
If you are looking for a community with an open friendly atmosphere to chat and share ideas, then Stack Exchange is not really it. Not that people are not friendly here, just that the atmosphere is more business like and focused around content production. It has more in common with Wikipedia than an machine learning forum. On the plus side the focus on getting good questions and answers is a great learning resource. On the minus side, it's not a place to make friends.
I suggest, in addition to finding a niche for yourself in Stack Exchange, that you look at other places with a more social atmosphere. Depending on what learning challenges you set yourself, you might find taking part in a free MOOC course will be fun - one classic machine learning course is Andrew Ng's Machine Learning on Coursera and a new cohort will be starting soon. The forums there are chatty, and also taking that course might inspire you to ask specific questions here to help you learn.