The Columbia 30000 series was used very sparingly as far as I can tell, which makes dating it without access to actual Columbia files very dicey. As always, I hope those better versed than I am at dating the majors will manage to find a true release date.
That is indeed a perfect example, Whyperion. Yes, the user who entered that date at https://www.discogs.com/release/7803600-Times-Winding-Up-Anyhow/history#latest claimed to have found it in Henry König's Labellisten. But when you search "Anyhow" by title, Columbia 37477, Columbia 30044 and the original release on Okeh 6238 are listed, but no publication date (Vö-Dat.=Veröffentlichungsdatum) is given.
Moreover, Henry König's lists frequently rely on sources that are themselves not always reliable, such as The Online Discographical Project at http://www.78discography.com/. For example, when you check out Okeh 6238 on http://www.78discography.com/OK6000.htm, it offers the wrong recording date for "Time's Winding Up" (mx. 30284). Moreover, the Online Discographical Project does not provide release dates anyway.
That said, for "Anyhow", König's source is Deezer (De). Not sure how reliable that is as a discographical source. It clearly doesn't provide release dates for any of the three different issues of this record, though, and neither does Henry König.
I don't think, König has ever bothered to look at Columbia sales catalogs, not to mention internal Columbia documents. That's the kind of labor that informs the work of real discographers like William Brown's "The Columbia Records (U.S.) 35000 / 40000 Series Popular Singles Discography 1939-1974," New York, NY, 1996. Brown got his information from the Columbia files in the Sony Archives. König's software, as impressive as it is, simply compiles information from internet sources.
Using Discogs release dates is like walking on thin ice. This time you were lucky, I guess. But in general, I would always make sure that the dates there are properly sourced and bolstered by actual evidence.