It's pretty simple really: there's no particular reason to imagine they were real at all. They lived on an island (Scheria) that doesn't exist, there's no evidence of such a people outside the context of the Odysseus story, they supposedly had ships that sailed themselves, they enjoyed ridiculous wealth and plenty without having to put any effort into it, they used to live next door to the Cyclopes, and their king is a grandson of Poseidon, a great-grandson of a Giant, and married to his own niece. There is no basis for seeing anything historical in them.

However, it was certainly fashionable in antiquity, as in the present day, to try to equate the fantastic locations of Odysseus' wanderings with real places: most of the locations mentioned in Odyssey books 9-12 were allocated real locations in Sicily and up the west coast of southern Italy. Scheria was routinely identified with the real Kérkira (or Corfu; ancient Korkyra). This identification first appears in Thucydides (end of the 5th century BCE), as a bit of Korkyran propaganda about their naval supremacy, in the context of a conflict with Corinth in the 430s BCE. The 3rd-century-BCE poets Kallimachos and Apollonios followed this identification. Among ancient sources, only Strabo resists it. So this identification serves as the consensus on what historical-era Greeks believed about the Phaeacians, regardless of the actual non-existence of that people.

It's not strictly speaking ruled out that the mythical locations described in the Odyssey had always been imagined as being the real locations that are quoted in later sources (Scheria = Kérkira, Cyclops island = Sicily, Sirens = rocks south of Sorrento, Circe = promontory in Latium, etc.). However, the vagueness of the geographical indications in the epic itself doesn't exactly support that idea either; in particular, the inaccuracy of the description of Ithake/Itháki (Od. 9.21-7) tends to suggest poor geography rather than intentional vagueness. And a reference in the fragments of a lost poem, the Naupactia (6th century BCE?; fr. 9 West = Pausanias 2.3.9) refers to Korkyra and not Scheria in a mythological context, implying that the equation post-dates that poem. So the first equation of Korkyra with the mythical Scheria would be later than the Naupactia (ca. 600 BCE, at the very earliest) and earlier than the conflict over Epidamnos (ca. 433 BCE).