I hadn't considered it could be seen as keyword cramming, but it makes sense. Would you say this even applies to pages marked with "no index, no follow"?
That's a good point, blocking the URL from being included in "natural" results with a meta robots "noindex" directive, should certainly prevent the document from being devalued by a "keyword stuffing/cramming" algorithm, and prevent it having any "knock-on" effects. Especially if you are using "hard sell" titles and copy for a paid campaign "landing page" URL such as Adwords, where agressive or OTT 'on-page' optimisation" and a keyworded URI will improve your quality score and you do not want that document to become part of the site.
Having a "nofollow" directive as well is debatable as that URL might be valuable as an "entry point" into the site for URL discovery, and therefore allow "deeper" URls to be indexed.