Oxyaenids lost flexibility in their limbs as they got larger, this was first noted by Osborn back in 1900, Patriofelis has much more rigid forelimbs than Oxyaena and flatter almost hoof-like claws so it's not an stretch to think Sarkastodon went further along that path of limbs useless for prey capture.
CMNH 63802 is from the Alaskan peninsula based on it being referred to U. arctos gyas in older publications, but supposedly peninsular and Kodiak island brown bears are genetically indistinguishable.
By convention Y is the dependent variable, the result of the equation, in that publication X is the weight instead of Y because the equations are meant to find out how long the bones "should be" at a given weight, so it is reversed.
Yeah, you are supposed to get mean values from your results, this is so proportional differences among taxa can be "averaged out" as much as possible, so, just because you are getting higher numbers with average cave bear humeri than the huge brown bear femur, doesn't mean that average cave bears were larger than it, if that huge brown bear had similar proportions to 63802, then its own humerus must have been almost 490mm long and getting an estimate of almost 800kg. Still, I do have my caveats with using those equations by Christiansen, I have good reasons to believe his largest individual was obese and has odd proportions, being that the reason why the humerus estimates are so high absolutely and relative to the femur ones.
That Italian humerus, that's 30% larger than second place and it wasn't even acknowledged, which makes me think scale bars are to blame again.
As for that brown bear skeleton, it resembles Mauricio Anton's art, it might be his, as far as I can verify the proportions are ok too. What do you mean by superfluous? it doesn't seem particularly big?