The main task for witnesses is to produce blocks. Without blocks being produced chain does not work. And yes, missed block, even from top witness, is not that big deal, but many missed blocks lower the security of the chain. It is not a coincidence that there are mechanisms to turn off the chain when too many blocks are missed (from many witnesses). There are other essential tasks for witnesses, like publishing feed (again, feed turns off if witnesses don't refresh it regularly).

Another very important task is to know what they are running. Changes in core code influence how the chain works. Witnesses take responsibility for that, so they should be aware of all the changes that are introduced with hardforks. Yes, it requires very different set of skills compared to node maintenance. They might need to hire specialists or spend countless hours tracking the changes, and the task is not even easy for people that are actively developing that code. At the very least witnesses must take part in testing.

Sometimes unexpected things happen. Top witnesses are similar to upper management in the corporation. Track record is what makes them different from "young blood" that would want to take their place. When someone took part in leading the chain through many hardships, it adds to their value as a witness.

As for secondary parameters such as HBD APR, I'd rather see mechanisms changed to something that would move the power away from witnesses-politicians to market forces. However until such changes are made, it is also an important factor.

For everything else - no. We could expect those things from witnesses if Hive price was in high double or triple digits. In such environment being top witness would mean you employ a diverse team of PR specialists, admins and software engineers, product managers, sales people etc.