Okay, all volatile gasses are gone now? So maybe I can light a candle and try to shed some light on this (or obscure it even more).
Jheem is of course right. The ge- prefix can be combined with nouns and I was not thinking along those lines and I should have.
As far as Dutch is concerned, linguists don’t talk about different prefixes, but distinguish different uses of the same prefix. The WNT distinguishes five different categories (with many sub-categories). But if you say that’s nit-picking, I’m not offended.
Point is (I think) that Foolscap proposed a Dutch word ‘geboedel’ as origin for ‘caboodle’. I like that idea and his reasoning very much because the step from geboedel > caboodle is smaller than from boedel > caboodle. My concern was and is that I can find no proof of that ‘geboedel’ ever existed. Moreover, all the online sources mention ‘boedel’ as the source.
The entry in the WNT mentions under ge- as category 1bβ that words formed with this prefix can indicate:
Collections of different objects, such as geboomte, gebeente* ,...
*from ‘boom’ (tree) and ‘been’ (bone) respectively.
I think that one comes closest to what our ‘geboedel’ would be, even though (as confirmed by Van Dale) such words are usually formed by adding a suffix -t(e). In that case you might even expect ‘geboedelte’. But that assumption is not getting us any closer to an answer.