Nevertheless of the comment about own research, in the context of PDF, "flattening" has several meanings. In general, what it does is consolidate everything into the imaging level, creating a simple PDF.
Main meanings of "flattening":
• In the context of forms and annotations in general: Rendering the annotations and/or form fields according to their visibility properties on the imaging level, and removing them. With that, you remove active elements, and form fields are no longer fillable, and their contents is no longer accessible.
• In the context of prepress/printing: Rendering sections where transparent objects are covering other contents. Only output devices which have a RIP understanding PDF 1.4 (or even 1.5) can render transparencies properly. When PostScript is involved, or the RIP can not understand PDF 1.4 or newer, transparent elements will be rendered opaque. The transparency flattening function (in Acrobat X and newer) will render (essentially rasterize) such areas according to the settings in Acrobat. Of course, you will lose the vector information, but at least, the page can be properly printed.
• In the context of OCGs (Optional Content Groups, also nicknamed as "layers"): transferring all elements of visible OCGs to the imaging level, and discarding all OCGs.
That said, unless you know the context, it can be difficult to come up with the exact answers.