So first error I noticed
REAGENT:A:xx
xx being the number of things you want
So we're clear, this error is present in the EXTRACT_FROM_PURIFIED_LUCRE reaction's tool reagent entry.
[PRODUCT:100:1:GLOB:NONE:GET_MATERIAL_FROM_REAGENT:lucre:LUCRE_MAT]
[PRODUCT_TO_CONTAINER:flask]
:GET_MATERIAL_FROM_REAGENT:xx:
xx being the reagent letter of lucre
This isn't actually a problem; "lucre" is, in fact, an acceptable reagent "letter." Look at the vanilla MAKE_SHARP_ROCK reaction in reaction_other; it uses "tool stone" and "hammer stone" as reagent "letters," including getting the sharpened rock's material from the "tool stone" reagent. Reagent "letter" should more appropriately be called reagent "string," as any unique series of alphanumeric characters works in those slots.
I don't think LUCRE_MAT is necessary.
This I am unsure of, but going from the soap reactions, it's probably best to have it. Given that globs can be many different things, making sure the game knows they're the right material type is a good idea. It should just ignore it if it isn't necessary.
PRODUCT_TO_CONTAINER:xx
xx being the reagent letter of the flask
Again, "flask" is an acceptable reagent string, so this isn't a problem.
And (according to the wiki) put [PRESERVE_REAGENT] after the flask reaction.
Where, specifically, do you see a flask reagent that doesn't include [PRESERVE_REAGENT]? I'm not seeing any.
As for the specific problem spots the op was worried about, I see no problems with them. The only issues I can think of are that globs might not be capable of being stored in flasks, that flasks might not be capable of storing things at all, and that the game might not recognize the lucre material. Based off of the soap and plaster powder reactions, though, it looks fine.