I would question this result in almost every meaningful demographic. I've sat and watched people fail to discover non-standard navigation methods countless times in testing of products across 12-18 and 20-60 (students and teachers separately) every time one is provided as part of a supplier product. Unless there was some kind of conscious or subconscious prompting going on I would doubt that anyone would genuinely favor it.
You can easily, to the detriment of results, create an environment where people say they like something when really they like the impact the the thing had on them the first time ("oh, cool") when really they don't prefer it at all but the record only reflects their initial impression.
Demographics are more varied than that though - they also vary by Country. Take Denmark, the birthplace of several of the major players in flip catalogs - online flip based catalogs are so common that every internet going user knows what they are, and how to interact with them.
With such a common use of them, we can start looking at how to optimize the catalog format itself, rather than stick with the scrolling vs. catalog format discussion.
You can easily, to the detriment of results, create an environment where people say they like something when really they like the impact the the thing had on them the first time ("oh, cool") when really they don't prefer it at all but the record only reflects their initial impression.