Longer term project: a few details

This idea originally was slated to be detail #... but as I kept thinking about it, it occurred to me that it fits with the isolating language I have been slowly compiling ideas for. I will, however, describe the idea as it developed.

each, every: formed as two opposite nouns or adjectives along these lines -

man and woman }
beast and fowl     }   NP
child and adult     }
large and small     }

E.g. every able worker → young and old able worker | big and small able worker
Before it occurred to me that this could work with the isolating language, I considered case marking and how it interacts with this. The following options occurred:

• These phrases, when used as determiners and pronouns, use the same case morphemes - and the same case distinctions - that pronouns normally use? Or alternatively, a simplified case system or one that merges cases in a way that differs from nouns in some other way.
• The case of the embedded noun phrase should be somewhat limited - partitive or genitive or nominative or somesuch, but maybe exceptionally mark accusative case when the phrase is the direct object of a verb?
• When used as a pronoun, the phrase is more often oblique than a direct argument of the verb, thus distinguishing it from sentences where these noun phrases are coordinated due to the speaker actually speaking of such combinations.

At this point, it dawned on me that maybe one could use articles to distinguish 'actual' use of nouns and this kind of use - omitting the articles whenever the use is pronoun-/determiner-like. At that point, this makes sense in an isolating language as well. It also dawned on me that the more typical determiner/pronoun 'all' probably also could originate with this construction, and an earlier 'all' could have been grammaticalized into some kind of aspect marker particle, but also still persist in some idioms. 

Thus, the language will have articles - probably using some kind of noun class system (with some overlap between classes), as well as a very free way of forming 'all, each, every'. Finally, fossilized use of an earlier 'all', as well as use of that earlier pronoun as an aspect marker. Oh, and the 'all, each, every'-construction should probably be more likely to occur in somewhat oblique positions than as direct objects.