You are in a tough situation. I also went to a small liberal arts college in the US. It had about 2000 students then, but has grown. We had good and bad professors but the good ones made up for the others. Note that in the US, grad applications rely fairly heavily on letters of recommendation, and I don't know if that will be an issue for you.
I normally suggest that students go directly from undergraduate to doctoral programs, partly because that is generally the norm (US) and partly because doctoral study is funded through teaching assistantships. Masters study rarely is funded.
Perhaps you should consider applying for both doctoral study and masters study (different institutions) simultaneously. A course based masters would make "hole filling" possible, though the funding would be from you. Your situation is also unusual (US) to be "complete" in the third year. Usually it takes four. And, it is also usual here to have the undergrad program very broad based beyond any major subject. That is unlike, say, Germany, where it is very focused on one subject. If your program is unusual in other ways then it might cause additional problems.
Math is very difficult to study on your own as a novice. It is hard to gain insight without guidance (and hard work). In reading your post, it is also hard to parse the issue with your advisor. You and I might have the same issue that you and your advisor do, but that is because when you would come to me for guidance and asking questions, I would try hard to give you minimal answers rather than complete ones. I might point out where you went wrong, rather than how to get right. I might point out a block. I might send you off for a reading. The purpose of such a style and philosophy is that the insight comes from working it out yourself, not from hearing or reading the answers of others. Math is a (very) active sport. It requires changing your brain physically, creating new pathways. Reading and listening are never enough until you gain the insight. Perhaps that is the real issue with the advisor and you are interpreting it as "not caring". Not necessarily, but a possibility.
To enter grad school you will need good letters of recommendation, one almost certainly from this advisor. Do what you can to cultivate a good relationship.
When it comes to applying for grad study, cast a wide net. There are many good schools with a "ranking" around 50 or so. Don't apply to just top schools. One hint, is to determine if some recent math graduate of your school has gone somewhere and done very well. Consider applying to the same institution.
As to how to do well in courses with minimal guidance, it is very difficult. Find textbooks with lots of exercises and do as many of them as you can. Some books have answers given for alternate exercises so that you can compare your solution to the one provided. Don't just read the solutions, though, as it is nearly useless. And, solve a lot of the problems. Learning depends on both repetition and feedback, though you indicate the feedback from the instructor is weak. Online courses that don't provide feedback is too passive for real, insightful, learning.
If permitted, working groups of students can be helpful, provided that you adhere to the rules set. But you can study together and give feedback to one another in most cases.
I'll keep thinking about this for a bit. Maybe I can come up with something.
