What do you mean, different instances of text? Do you mean different types of text that get rendered differently? Or does this apply also when you work with a large block of text that's all rendered a single way?
Basically, each block tag (h1, h2, h3, p, li, etc.) would create a new instance of sIFR, thus creating a new flash object on the page. For example, you can't use just one flash object for a long stream of text, if that stream of text is divided into paragraphs, because sIFR cannot internally give you the necessary margins or tabbing required for marking paragraphs.
The only case where you would have one flash object is when you have a single column of text which has the same font, same leading, and no margins throughout. Unfortunately, that won't happen very often.
The only case where you would have one flash object is when you have a single column of text which has the same font, same leading, and no margins throughout. Unfortunately, that won't happen very often.