Recently, while reading M.L. West's book Textual Criticism and Editorial Technique, 1973, I came across this passage on p.11 about the usefulness of imitations:
Imitations and parodies are occasionally of use, especially in the case of verse texts. For example, in Iliad 1.4-5, where the main tradition gives αὐτοὺς δὲ ἑλώρια τεῦχε κύνεσσιν οἰωνοῖσί τε πᾶσι, Athenaeus records that the pioneer of Alexandrian scholarship, Zenodotus, read the more forceful δαῖτα instead of πᾶσι: it has been observed that this is supported by the echo in Aeschylus, Suppl. 800f. κυσὶν δ᾽ ἔπειθ᾽ ἕλωρα κἀπιχωρίοις | ὄρνισι δεῖπνον. (But the imitation could not have been used to infer a reading δαῖτα if its existence had not been recorded.)This left me with the impression that West would read δαῖτα in his edition, but when I saw his edition shortly after that, I was surprised to see that he actually read πᾶσι. A friend was kind enough to send me West's note on this passage from his Text and Transmission, 2001:
His [Zenotodus] δαῖτα in 5 was apparently a variant familiar to Aeschylus and the model for his κυσὶν δ᾽ ἔπειθ᾽ ἕλωρα κἀπιχωρίοις | ὄρνισι δεῖπνον at Supp. 800f.* But that does not make it the original reading; πᾶσι is good idiom, cf. Ar. Av. 1117, and Soph. Aj. 830 has ῥιφθῶ κυσὶν πρόβλητος οἰωνοῖς θ᾽ ἕλωρ. Against δαῖτα Athenaeus cites (from Aristarchus) the argument that Homer uses the term only of human banquets, but this is not decisive, cf. Ω 43 (Leaf), and δεῖπνον in Β 383. Zenodotus' reading δηιοι for τ᾽ ἄλλοι at Ι 594 is of similar character--more colourful, but secondary.I still wonder what reading he would have used in 1973, but it is a good reminder that the "more forceful" or "more colorful" reading is not necessarily the original one.
[*Other similar passages such as Eur. Hec. 1077, Ion 505, and Tim. Pers. 138 (cited by R. Renehan, AJP 100 , 473f.) may derive from Aeschylus and are not necessarily independent witnesses.]