yes you should almost as always want to use mechanical for landscapes. No reason not to especially with EFCS to avoid shutter shake.
So the measurements given on photonstophotos is photographic dynamic range (PDR).
The website is maintained by a professor at Harvard (Bill Claff), his explanation of PDR vs. DXO is given here:
Expert news, reviews and videos of the latest digital cameras, lenses, accessories, and phones. Get answers to your questions in our photography forums.
www.dpreview.com
His measurements are generally more reliable and accurate (I think with 5-10% margin according to Bill) than DXO. The error is kinda unavoidable because he builds his graphs from samples provided to him by various people (he doesn't and cannot afford all those bodies). You could contribute too if you wanted, its not hard and you can also do his measurements yourself. He will give you exact instruction on how to do both but being humans and all there will some minor errors here and there (in fact the instruction might now be on his website). But DXO is sometime just wrong, for example it doesn't even show the dual gain in dynamic range on some of the sensors. Plus there are many inconsistencies on DXO, for example the dynamic range of A99II is lower than that of A7RII on DXO even though Sony explicitly stated both bodies basically used the same sensor and processor.
In fact you can even contact Bill via. email or dpreview and he'll reply back to you clearing any doubt or concerns you have with measurements. So there is transparency in everything he does as supposed to DXO which is a blackbox.
A9's dynamic range was never great which is why it never interested me and still doesn't. Couldn't care less for it. If you want to see a dynamic range even worst have a look at D5 or D6. These bodies aren't built with high dynamic range in mind but for performing really well at high ISO for sport-togs. They do that pretty well. Personally I never shoot above ISO6400, only like 5-10% of my pics are between ISO3200-6400. So I only care about low ISO dynamic range for most bodies.
Based on his margin of errors, I wouldn't say A7RIV is better or worst than R5 (going by graphs alone that is). BUT one thing to note about R5 is that they are using some kind of baked in noise reduction at lower ISOs to "artificially" boost the dynamic range (denoted with the reverse triangles in the photonstophotos DPR graphs). So they are likely to have slightly less detail and perform not as well as other bodies with equal dynamic range and resolution like the Z7 or A7RIII/IV.
you can see the effect of that compared here on dpreview:
Canon's EOS R5 is one of the best mirrorless cameras on the market today. But even with its 45MP sensor, 20fps bursts and 8K video, its not without its flaws. Find out all the ups and downs of Canon's latest mirrorless flagship right here.
www.dpreview.com
I just realised I spend far too much time reading these things