A different direction than the others....
Cam lift, valve springs and locks etc. Is there a possibility that you are running a valve train combo that doesn't have enough clearance and when the valves open they are binding and putting more downward force on the cam. I have seen similar issues with guys putting big cams in engines that don't have the valvetrain to handle it and then if the lifters were adjusted too tight it can exacerbate the problem.
It could also be rockers binding at full lift as well.
It looks like the majority of the damage on the bearings is on the bottom of the bearings? If so I would think its a valve train binding issue.
Thanks to both of you for this suggestion of binding valve train. This would make sense for how the bearings failed, as all look to have failed exactly at the highest pressure point and "swept" molten babbit clockwise. That said, my heads are Edelbrock 5023 with a Comp Cams 256H cam. Lift is "only" .477 I/.484 E. When I measured for pushrod length I used a lifter I converted to be solid (stacked washers in to replace the spring). I ended up with about .125 longer than stock at 6.95". The cam lift I have is well within the cylinder head specs where they allow up to .55" lift.I will double check once the bearings come out, but I can say I am positive the cam bearings went in properly form back to front. I appreciate the reply and suggestion. I actually think it would be hard to put them in wrong as if you put one in the wrong place, as you would end up with at least one that is too loose to press. I believe each bearing is .015 smaller front to back.
I do see that Durabond offers 3 types of bearings and I have the lowest model (OE replacement). Still with my set-up, yes not stock, but still a mild street build.
Maybe this is my second mistake: 1) not measuring cam bearing bore and cam journals for clearance, 2) not specifically checking the valve train for binding. It was not on my risk radar as I was well within the specs of the heads.