I have comments specific to each course at the respective course sites. Here I have a few general comments to make.
Very roughly, 90% or better is an A, 80% to 90% is a B, and so on. I don't grade by points, though, so those numbers indicate a level of precision that simply doesn't exist. Rather, I evaluate the sum total of the work for any given assignment and then assign a letter grade to that evaluation.
I do give plus and minus grades. I mainly prefer A, B, C, but if I feel ambivalent about a grade, I'll reflect that by adding a plus or a minus to the letter.
The whole notion of "late" and "on time" in an asynchronous course gets a little fuzzy. These concepts make sense only in a real-time course in which there is a class meeting on a specific day and time; then, "on time" means "handed in physically during the class meeting time" while "late" means "handed in after the class meeting."
Our classes don't meet. You can submit an assignment at any time up to midnight (Mountain Time USA) of due date. If you chose to turn the assignment in on the due date, rather than some time earlier, and then you got sick, I'm sorry but I'm inflexible. The wise student will understand that it will be better to turn assignments in earlier rather than later.
Late assignments will be graded down. How much down will depend on how late, but I reserve the right to make subjective judgments in that regard. Don't worry, my subjective judgments are usually in favor of the student.
You can always turn an assignment in early.
In light of the above, it therefore follows that I do not give makeup work.
I do not give extra credit assignments. I put a good deal of consideration to planning a course's assignments and in balancing the relative weights of each. Extra credit assignments unbalance that to the detriment of the other students.
Moreover, I object philosophically to the notion of extra credit because it is based solely on quantity. Extra work that is mediocre is still mediocre and still earns a "C" grade. Doing more of it doesn't change the result.