In the question: Lengths of review process of mathematical journals a nice question was asked and answered: how to compare journals in terms of time for the review processing. The article cited in the answer is quite nice http://www.ams.org/notices/201310/rnoti-p1390.pdf but is from 2013. Seven years after, are there some similar studies which can tell what kind of backlogs/ times needed for mathematical journals to accept / reject articles? A young mathematician asked me this today and I gave him the link, but a more recent one would certainly be more helpful.
The American Mathematical Society provides regular (annual, I believe) on journal backlog times. The latest can be found here: https://www.ams.org/journals/notices/201910/rnoti-p1713.pdf Depending on your subfield it may not contain data on all relevant journals, but it is quite comprehensive. The reports are published in the November issue of the Notices of the AMS, which is also available online.
The data in the report is based on information provided by the journals itself.
A general source for many, many domains beyond just mathematics is SciRev, a site where authors submit reviews of their experience with the peer-review process at many (hundreds or thousands?) of journals across many disciplines. If you click on the "All reviews" tab, you can then search for journals by name or discipline; it includes many subfields of mathematics.
For me, the most useful feature of SciRev is that I use it to track my personal peer-review process, and then when a final decision is received, I can conveniently submit a review of my experience with that journal.
However, the downside of a site like this is that it is entirely self-reported, so without a large number of submissions for a given journal, you cannot consider the averages reported there to be scientifically valid. That said, it probably has reports for more journals than a scientific study like the one you originally mentioned, so it can at least give an indication of delay times.