Lifjeld and Bjerke have apparently shown, that flammea and cabaret do not interbreed in nature and, consequently, do not produce hybrids in nature, even though it is highly likely that they could, if they wanted. Under these circumstances it is irrelevant in the first place, whether hybrids are fertile or not, or have lower fitness. Or have Lifjeld and Bjerke's results been disproven or contradicted by others? I mean, have hybrids been found in nature in the meantime? Any references? I'm not up to date on Redpoll research, admittedly.
Based on Lifjeld and Bjerke alone (!), both taxa must undoubtedly be considered as biological species: they are diagnosably distinct and do not share a common gene pool. Accepting that their morphological, behavioural, vocal etc. differences have a genetic basis, there must be molecular differences, even though these may not yet have been localized in the genome, let alone quantified. The taxa are not fading one into the other through clinal variation. There's not even a localized hybrid zone (or is there elsewhere?). Hence they had separate evolutionary histories, long enough to account for those differences, and will very probably - due to their assortative mating - have separate evolutionary fates.
It is the taxonomist's job to describe biological diversity as it actually exists in nature, not what could be made of it under captive conditions. Otherwise we would have to lump Canary, Red Siskin and Hooded Siskin as well.
Rainer