OK, sorry I got as little hot when you asked for the 5/8 link, because I know you know we are tossing around ideas and possibilities, not proven designs.
Perhaps, as you say, a 5/8 director wouldn't work as a director but more like a reflector based upon resonant length, but what I'm wondering about is whether or not having this newer design, a center fed 5/8 dipole as per the Gainmaster might not effect an extended 5/8 field causing/forcing the parasitic elements to re-radiate and reflect/direct, since this type of extended current distribution hasn't been tried before as a driven element.
Since the Gainmaster stretches the usual maximum 1/2 wave current node out to 5/8 long, perhaps when excited by this extended current distribution driven element with those unique current phase angles, the parasitic elements might resonate differently than they would normally and mirror the Gainmaster current distribution?
Or perhaps the parasitic elements might need similar series/parallel phasing caps in order to work as such, if those elements are going to revert back to self-resonance even in the presence of such a unique excitation as the Gainmaster driver.
And as far as the gain factor, the Gainmaster 5/8 has all current in constructive phase unlike the conventional end fed 5/8 which loses about 40% of it's current distribution due to cancellation, and has only about 1.2db gain over a 1/2 wave.
In testing against my metal bottom fed 5/8 the Gainmaster outperforms it by at least 1.5-2db so add that to the 1.2dbd and you have more like 3dbd with relation to both the driven and beam parasitic element size, not just compared to a single element dipole.
I could see this being a 3dbY antenna where Y represents a Yagi of similar number elements
If you haven't read the Sirio Gainmaster page, it deserves a look.
