In this sense, Firth's turning language back on itself is just ('just' sounds a bit unfair here but is not intended evaluatively, only ideationally) the working through of the belated (OK, that one was evaluative...) Wittgensteinian (and already of course Peircean :-) ) realisation that we don't have any direct access to some 'out there' where we can attach labels. Got it, been there, bought the T-shirt.
 This misunderstands the Firth quote 'language turned back on itself'. It is not concerned with the realisation that Bateman attributes to Wittgenstein and Peirce. It simply means that linguistics involves using language to model language. Halliday (2002 : 384):
All systematic knowledge takes the form of ‘language about’ some phenomenon; but whereas the natural sciences are language about nature, and the social sciences are language about society, linguistics is language about language – “language turned back on itself ”, in Firth’s often quoted formulation.
 To be clear, just serves no ideational function at all; its interpersonal function is to enact 'counterexpectancy' of the type 'limiting'; that is, it limits what is to be expected (Halliday & Matthiessen 2014: 188).
In this sense
Firth's turning language back on itself
the working through of the belated Wittgensteinian (and already of course Peircean) realisation that we don't have any direct access to some 'out there' where we can attach labels
conjunctive Adjunct: matter
mood Adjunct: intensity: counterexpectancy: limiting
 On the SFL model, it is not a matter of having access to 'out there', because there are no meanings 'out there' to have access to. Instead, we transform our experience of 'out there' into meaning, distinguishing two orders: the 'material' (phenomena) and the semiotic (metaphenomena).
 Apparently, this is a variation on an idiom that expresses the speaker's complete familiarity with a situation, with overtones of cynicism or exhaustion. (This again is the language of bullying: 'I know this stuff backwards, so don't try to argue'.)