Of course I don't know how much meetings you have to schedule, but spending 1h/day for scheduling meetings as a full-time PA doesn't sound like a lot of time to me (? - would be interested to hear what others working in assistant roles think).  Scheduling meetings and doing it well so that the person you are supporting doesn't have to spend a lot of time on it themselves and don't worry about, is one of the things where a PA really add value. I think there is a limit on how 'efficient' you can be with scheduling meetings while still making it as easy as possible for the people involved in the meeting, so I wouldn't worry too much about spending some considerable amount of your time setting up meetings.

Besides using tools like calendly where appropriate the only other concrete piece of advice that comes to my mind is that is very useful to know the details of the preferences of the person (or people) you are scheduling meetings for, such as their prefered times for meeting, what kind of flexibility they have for other times of that's needed/the only option due to timezones, etc.

All else equal, higher salaries almost always  would lead to more candidates, but once you close to a level that is relatively competitive (eg compared to similar roles in other non-profits) and seems generally relatively fair compared to the rest of the organisation, I think other aspects of the role might become more important to candidates (eg the general work environment, how much freedom people get in the role, how much involved they can get in other aspects of the organisation if they want to, etc). 

In a (written) work test I would usually try to have a situation that is similar to something slightly challenging that would be encountered in the actual work (Something  that is based on an actual situation but adjusted so that it works well for a test environment is usually good.).  This allows to get at least some idea on how people would address  a problem at work in practice and I would try to incorporate to test for some of these skills:

• General problem-solving capability (by making the task open enough to allow for some creativity)
• Ability to follow instructions (by giving clear instructions on what needs to be done including relevant details)
• Ability to structure thoughts and organise things (by asking eg to create a high-level timeline for some event planning or other activity)
• Attention to detail (by a) potential documents the candidate creates and b) including bits in the tasks that get easily missed without a strong attention to detail) 
• (Written) communication skills (by asking eg to write a reply to an email as part of the task)

What exactly you would do would of course depend on the specific characteristics of the role (eg if the role involves a strong Finance component you would probably add something specifically relevant to that eg dealing with spreadsheets). 

For these kind of generalist skills, a work test in my experience works relatively well in separating very good from merely 'okay' candidates (And usually quite easily identifies bad candidates.)

There are other skills that are very well tested in interviews or similar situations and the combination of a work test and an interview has in my experience worked to judge candidates for (junior) ops roles.