I don't think you can just "by definition" reality like that. Sure, the non-reactive strain won't be called a symbiotic partner, by definition. But that doesn't poof the bacteria out of existence.

That said, I think you're correct about not being so concerned. If the symbiotic relationship was so exploitable, it'd probably be a great evolutionary advantage to do so. The signal being a sugar itself, as you note, doesn't surprise me.