Blass’s discussion slightly earlier [§56.1 – 1898, MacMillan, trans. H. St. J. Thackeray] affirms, perhaps slightly more explicitly, that the Greek tenses express time relations “absolutely, i.e. with reference to the stand-point of the speaker or narrator, and not relatively, i.e. with reference to something else that occurs in the speech or narrative” (italics added).
Blass’s subsequent discussion of the present in direct and indirect discourse (§56.9) also seems to illustrate your point about his view of the Greek tenses: “In the N.T. the use of oratorio obliqua is certainly not favored, and that of oratorio recta predominates; but it is noteworthy that subordinate sentences after verbs of perception and belief are assimilated to oratorio recta, and the tenses therefore have a relative meaning” (italics added). Blass then cites Matt 2:22 and John 6:24. In these cases, then, the tense in the quoted discourse is relative with respect to the writer who reports the discourse and absolute (see §56.1) with respect to the person(s) reported as constructing the discourse originally.
So, to change the terminology set, it seems that, even if the tenses in the original discourse were “objective” [i.e., that they were directly tied to the reality described by the person(s) who originally used these tenses], the quoted forms themselves properly describe only the original user(s) “subjective” perspective, which may be quite distinguishable from the perspective of the person who later quotes this original material (e.g., Matt 5:43–45; 11:18). If this reading of Blass is correct, then your point about his agreement that tense forms of themselves do not necessarily correlate with reality becomes even more sure.