That's exactly what it shows; criteria was not met.  Possibly because I have some document classes set up to limit the number of pages in the Last Page Identification section.  I could remove that limitation and maybe that will get it working...

Last page identification could be a factor. If your Last Page Identification is cutting the document off prematurely, then I believe the remaining pages would get processed as the start of a new document and may not meet any of the identification criteria.

You might see mention of that in the output as well if you scroll through and check the Messages for all those steps.

Removing the last page identification resolves the issue, but presents another question.  If the second page of a document is something that I want to remove, it seems to me that QF should treat the second page removal request as something that should be done immediately and not run it through every doc class first.  Logically, at least to me, if the first page identification has identified the document, then the page removal task in the Last Page Identification should remove the requested page. 

I think some clarification is needed regarding first/last page identification.

Removing a page doesn't inherently mean that every page will be run through all of your document classifications. The reason your blank pages are being run through the classification is because they are being treated as a new document.

Last page identification means exactly that, the last page of the document, so if you set it up so it "completes" a document at the blank page, everything after that becomes a new document, hence the classification checks.

The logical reasoning behind this is that Quick Fields can be used to parse individual documents out of a consolidated file. For example, if you pull in a PDF that consists of multiple scanned invoices, the first/last page identification allows you to parse them out into separate documents when they're processed.

However, in your case, you just want to remove the blank items from the document, in which case you would want that to occur within the classification activities rather than splitting them off into a new document.

Instead of creating a new document by triggering Last Page Identification, try adding your page removal to the Page Processing or Post-Processing activity

Add the Page Removal activity, then set it to the parameters you want, which I believe you said was based on file size.

That's what I'm now doing for the blank pages and it works fine.  I was trying to do something similar for some other documents that have a second page that is not necessary but has text (such as page 2 of a W4), so I specified removing page 2 in the Post-Processing page removal task.  Instead of removing page 2, however, it just continues on to the following doc classes trying to classify the page.

Something else had to be going on as a result of Last Page Identification rather than the page removal.

I have several QF profiles that remove pages and I've never had it try to process the removed page as a separate classification.

Post-Processing is the very last thing to happen, so it sounds your "extra" page is getting split off and processed separately before it even reaches the "Page Removal" task.

That has to be the case because Page Removal is literally just deleting the page so it would never trigger a new document classification.