These splits are tentative. The notes to each entry emphasise that, especially the statement 'Redpoll taxonomy is unsettled'...
An equally persuasive case can be made for lumping, given that Mason and Taylor 2015 found little to no genetic separation at species level between any Redpoll populations, whether previously considered separable at species level or not.
Either way, there are too many vagaries that don't at all fit with either approach, which means that further work may shed some light one way or the other. Note also that Amouret et al 2016 made a case for morphological separation in most Redpoll taxa, except for the Iceland taxon which they were researching!
Amouret, J, GT Hallgrimsson, Y Kolbeinsson and S Pálsson. 2016. Morphological differentiation of Icelandic Redpolls, Acanthis flammea islandica. Bird Study 63(1): 37-45
Mason, NA and SA Taylor. 2015. Differentially expressed genes match bill morphology and plumage despite largely undifferentiated genomes in a Holarctic songbird. Mol. Ecol.doi: 10.1111/mec.13140
Yep, true - but IOC currently accept them, and for the purposes of this list, "ours not to reason why, ours but to do and die"
(well, I hope not from this wretched virus, anyway!)