ok, so I will begin using fer- for what Koebler has far- for (he does note that they are variants of one another). The faur- = "fore-" makes sense to me (I count this as a separate prefix altogether) but I wonder whether Gothic may have also experienced some merging into fra-, and that fair- was a relic.
Will this then work: fra- = (far) away (from), off; fur- = fore-, before; fer- = completely, fully, up?
Also, if you consider the regular operations of sound laws, Germanic unstressed -er- remains and doesn't become -ir- (while unstressed -e- otherwise does become -i-). But in most words -er- later appears as -ar-, especially in West Germanic. So it's possible for West Germanic *far- to derive from earlier *fer-. Gothic *fair- can come from either *fer- or *fir-. The appearance of *fir- in West Germanic doesn't have to be significant, because *ta also appears as *ti and *ga- as *gi-, so it seems that prefixes involving *a generally became merged with *i fairly early already. The same could have happened with *fur- too.