Horsfall on his own again (with only one other commentator). Unsurprisingly, I had sided with Horsfall, but with McGill adding yet another vote to the majority...
spem si quam ascitis Aetolum habuistis in armis,
ponite. spes sibi quisque; sed haec quam angusta videtis.
cetera qua rerum iaceant perculsa ruina,
ante oculos interque manus sunt omnia vestras.
His complete note:
haec quam angusta There exists an almost total accord among commentators—or at least among those who deign to consider the words—to take these words as fem. sing., in agreement with spes. Perret, however, prefers (as I had myself suspected) neut. plur., haec angusta parallel to cetera . . . perculsa and thematically comparable to 304, 311. Spes angusta seems not to be a standard description (Hey, TLL 2.63.59) and the issue is initially not perspicuous, though a first glimmer of sense appears when we realise that haec shows Lat. pointing; the recent battles have left Latin manpower straitened, angusta (cf. res angusta domi!). V. has indeed some striking changes of subject (7.211, 241), but does not write to set puzzles, no more than Hor. does. Explicit indications present in the text are provokingly abundant, but, it may have been thought, inconclusive (in that both sed and cetera point to changes of direction, the latter in evident contrast to haec). Perret understandably prints a comma at the end of 309 but the full stop in other edd. has no sacred authority; no more, though, does the colon after 310 ruina in Perret (a comma in other edd.). The eye that faces (or ear that hears) this text for the first time (or pretends to; cf. my remarks in Studi Tardoantichi 9 ( 1990 [in fact, 2000!]), 18ff. for this method applied at Hor.C. 4.4.13ff.) will naturally pause after both ponite and (though not necessarily for as long) quisque; thereafter an honest perplexity may be forgiven, indeed commended. However, given the evident balance (and one well enough attested: cf. Lucr.l.402f., Ov.Met.l3.957, Trist.4.3.27) of haec and cetera (reinforced as it is by the further balance angusta::perculsa), it is very much easier to supply sint with haec quam angusta and omnia could then just as naturally (if not rather more so) take up both haec and cetera as cetera alone. In that case, haec . . . angusta are perforce neut. plur..