Zwicky, in Exceptional Degree Markers, describes the
too big of a dog
expression as being confined to [some] American dialects (p 113; see also footnotes). He seems to criticise the usage, saying that 'Clearly, of is now something more than a mere preposition. It's a virus.' He links to Abney (who claims it is dialectic) and Radford (who comments
'Abney (1987: 324) notes that in non-standard forms of English we find
structures such as the following: (37)
a. too big of a house b. as nice of a man c. how long of a board'
), and mentions some 'mildly alarmed non-linguists'.
In Language Log: Bundling, he gives the following examples of and comments on 'intrusive of':
On to "intrusive" of. Here, many commenters bundle P + of
(in alongside/inside/off/out/outside of)
together the of that appears in
one variant of exceptional degree modification (the much-reviled too
big of a dog as an alternative to too big a dog), but the two
phenomena have nothing to do with one another beyond that of.
There's extensive discussion of the five P + of cases above in this
course handout of mine [qv]. For these, there’s a separate story for each
one (though some handbooks recommend against P + of in general): plain
out is extremely restricted; outside of is not colloquial (except in
one sense); off of is somewhat on the conversational side; etc. Off of
is the combination that gets the heaviest criticism, though I don't
think that on the evidence of actual use, it can be classified as
non-standard — on the colloquial side, but not non-standard.
Admittedly, this is 25-30 years old, but the descriptors 'dialectic', 'non-standard', 'mildly alarming' and even 'a virus' and 'much reviled' should perhaps urge care in using this construction.