If changes are made in a submission as a result of reviews, can the submitting person access and change the original submission?
No. One resubmits, creating, in effect, a fresh document. This avoids the garbling that sometimes comes with insertions and deletions. It does mean reentering some routine matter, so it's a good idea to keep a copy of the responses to the queries in your original submission.
If I have several independent items of the same general type, should I enter them as a single submission or as individual submissions?
Use your judgment-the site can handle either. A good way to decide is to put yourself in the place of a user, who initially sees only the titles of the submissions. Thus if the items group naturally ("Three demonstrations of scaling") lumping might be reasonable, otherwise ("Useful images for lectures") they might better be divided.
What if I come up with improvements the next time I offer my course?
Ahh-an advantage of such a site. You can just resubmit another version, even one amended by a single word, with an e-mail note to one of the editors of the appropriate section (see: Contacts), who can then check the change, post the new version and "unpublish" (!) the old one.
What if my submission does not fit comfortably into any single one of the categories of the site?
Do not worry much about the matter; note, if you want reassurance, the uncertain fit of most of what is already there. The categories (which may be changed) have been designed as generics that can work for other sites in the network as well.
What, perhaps after use, if I have ideas for improving material already posted on the site?
You might begin by contacting the original instigator by e-mail. The "unpublish" option allows changes, either as internal addenda or as reworkings. Ordinarily, nothing will be altered within a submission except with concurrence of its original author. Alternatively (and less desirably), you can send in a separate submission with a link (a URL) to the original.