Accent neutral, in this case, are voice sets that embody archetypes, without displaying any overt or obvious hallmarks of real-world national/cultural accents. Most video games attempt to get as close to possible to this with at least a few of their voice sets. Voice coaching for VAs is a thing precisely for that reason (amongst others, of course) - They'll take a set of lines or barks, listen through, note the sounds and inflections that have strong signifier factors for particular languages/dialects, and smooth them down on subsequent takes until they're mostly unnoticeable, and the voice coach can no longer say with any clarity what sort of an 'accent' the line has at all.
Oohhhh okay, thank you.
Usual fantasy RPG voices sound mostly British to me, still, but I didn't know that.