I have not had time to look into the matter thoroughly. However a quick
check of Stembergers “Introduction to Talmud and Midrash” the standard
handbook of modern scholarship of rabbinic literature suggests that R.
Zevieli has been rather selective in his citation of the scholarly
literature on the topic of megilat Antiochus. Stemberger writes:
proposes to date the work for linguistic reasons between the second and
fifth centuries, but most would assume the eighth or ninth century and
regard the language as a literary imitation of targum Onqelos.
(Similarly A. Kasher, along with other authors… considers… the text to
be… redacted in polemical reaction against the karaites, who rejected
this feast, this text could not have been composed before the second
half of the eighth century…)”
In short, Kaddari is at best, a
daas yachid on the issue, whose opinion has not been accepted. That
does not make him wrong, of course.
As for other sources for the
nes chanuka, there is only one, the famous “mai chanuka” baraita in
masseches Shabbos. A similar text appears in many version of megilat
taanit, but numerous scholars, most recently vered noam, in her edition
of megillat taanit, argue that this is a latter interpolation from the
[Talmud] Bavli and not an independent source. These scholars rely in part on the
fact that the Or Zarua cites megillat taanit with out any reference to
the nes pach hashemen.
Even if we accept that this baraita did
in fact originate in Eretz Yisrael around the third century, (not a
simple assumption given that neither the Yerushalmi nor any EY
midrashim, even those as late a pesikta rabbati, seem to have been
aware of this baraita) this still places it centuries after the time of
the hasmonean revolt and much latter than many other accounts of
origins of chanuka. Lets not evade the issue of the nes pach shemen,
it’s a problem.