There are two fairly easy ways to handle this:
1) Provide filters that will do an automatic, half-hearted job of markup.
Simple things like converting tabs (converting them into html3 tabs when
those are supported, converting them into strings of blanks for now) is
simple, and in fact the unix utility "expand" is used routinely for that.
Form feeds could be filtered out, changed to line breaks, or translated to
<hr>. With only a bit more work, form feeds could be used to divide a
plain-text document into muliple linked html documents, which is the
closest www equivalent to page breaks in scanned documents.
2) If even that is too much work, the documents can be served up by gopher
rather than http. This trivially preserves the structure that is present.
>I have the job of presenting official information to the public
>in stable text and in reports generated from databases.
Then presumably you have the skill to write the simple scripts suggested in
item (1) above.
>I have very basic needs and I think HTML should meet those needs
>without style sheets. There are things that do not belong on the same
>page. I need HTML to say something about pageing which I can expect
>the general browser that the public will be using to support.
As mentioned above, the closest equivalent is to produce one html file for
each page. (That's why they are often called "web pages.")
>I dont need a special tag if the standard says that browsers should
>have as an option in printing a skipp to a new page on <DIV>
>or on <DIV CLASS=something> or something of that kind.
>Is that cluttering up the standard?
Yes, "cluttering up the standard" is a good description.