I think it's safe to say that the key thing is that
most of the "math stuff" you write should be about the same size as the text stuff, so that the numbers aren't, say, twice as big as the letters you write
Subscripts and superscripts should be a little smaller, so that it's easy to see that $2^4$ is not $24$.
Capital vs lower-case letters for naming things in algebra are a matter of context. Many people, for instance, use $C$ to represent a constant, but a few, in other contexts, use $c$, or some other letter. Physicists seldom do, however, because $c$ (for them) means the speed if light.
If there's one overarching rule, it's that your writing should be so clear and unambiguous that you still know what it means when you look at it again two weeks later.
As for writing on "graph paper" --- some folks will put $1/3$ in a single cell, some will write $\frac{1}{3}$ so that the "_" is in the cell, the $1$ crosses into the cell above, and the $3$ crosses into the cell below, others might use three vertical cells for the whole thing. The advantage of the more compact forms is that you can put more on one page, which lets you look at a single page and grasp the "big picture" of whatever you're writing, which (in math) often occupies multiple pages, even though the notation is designed to be compact and expressive.
Mostly: don't sweat it. Use what works for you, and what you can still read later.