According to my research, this isn't true, at least as far as text products (books) are concerned. Here are the total individual text products published by Paizo, for 3.5, P1, Starfinder and P2, according to my charts:Note that Paizo's product schedule, as large is it is now, was far larger in the 2013-2017 era when PF1 was the 500lb gorilla.
2007: 4 (all 3.5)
2008: 24 (all 3.5)
2009: 28 (13 3.5, 15 P1)
2010: 37 (all P1)
2011: 41 (all P1)
2012: 40 (all P1)
2013: 41 (all P1)
2014: 41 (all P1)
2015: 40 (all P1)
2016: 41 (all P1)
2017: 38 (34 P1, 4 SF)
2018: 48 (37 P1, 11 SF)
2019: 46 (20 P1, 11 P2, 15 SF)
2020: 39 (1 P1, 23 P2, 15 SF)
The number actually went up in 2018-19; the 2020 numbers include only what has been published or announced, so may change.
So, yes: the number of Pathfinder 1 books published is "far larger" than the number of Pathfinder 2 books, but the number of Paizo books is roughly the same or more over the last few years.
This might explain the possible reduction in books in 2020; that, combined with them not supporting P1 anymore.Still , core rulebooks, which can be printed in the hundreds of thousands of copies over the course of a game's life is where the real money is in RPG publishing, so WotC focuses on those sales via retailers. Those same FLGS retailers are the ones who generate the real money selling hobby games: which is to say, selling Magic:TG for WotC. [All RPGs taken together are modest compared to MAGIC:TG. It also explains why WotC has been reticent to sell directly. WotC depends on those retailer relationships.]
The direct sales model allowed Paizo to dominate the RPG business from 2010-2016. But when 5e started to gain some real legs in the marketplace, by early 2017, Paizo's retail sales via FLGS had dropped significantly.
Players like direct sales and the free PDFs a subscription provides. FLGS store owners? Not so much.
Now if we compare a typical P1 year and 2020, which is the first full P2 year, we have 40ish vs. 23 books. The big difference in terms of what is being published are the Player Companions. 2020 sees 3 hardcover splats; P1 saw a bit over 3 per year, from 2011-18 (usually 3, 4 a couple times, and 2 once). Campaign setting books have reduced from about 9 a year (plus or minus), to 4 -- but hardcovers about twice the length. Both see about a dozen adventure path episodes. There are also about half as may modules. But the dig difference are the Player Companions; P1 usually saw 10-12 a year, which hasn't been replaced.