I think what's needed more than anything right now then, is a consensus of what makes current 6es maps good when creating potentially new maps by looking at how current maps are played. I don't know a huge amount about hammer although I've tried using it a couple of times before but it seems to me like the hardest problems to fix involve things around the 2nd points of maps and the connectors between points as it involves changing large parts of the map as everything there is connected to something else. Mids seem to be an area of really common complaint on maps. Lasts historically seem like they've been easier to fix.
I might detail some of the stuff that I feel about 2nds and connectors in another post, but I can talk about mids right away.
In the 5cp maps in etf2l's rotation there generally seem to be 2 kinds of mid: Mids that feel "large" and mids that feel "small".
"Large" mids all seem to have something in common- the point is on the floor with quite significant height advantage around it (e.g. Granary, Sunshine, Process). The footprint of the mid on the map isn't necessarily bigger than somewhere like badlands with the huge valley area but more of this is used because of the way the mid is played. On these mids, when playing aggressive, most teams seem to try and take a side and push across the point on this side together often going around the point rather than directly over it.
Maps with "small" mids tend to have the point raised rather than the surrounding area, often on a bridge (e.g. Badlands, Gullywash). Most teams play or jump onto the point and push across to the enemy team, or jump the enemy team's side and try and take it, rather than going "around" or under the point, so teams tend to end up playing closer to each other.
The only map that really doesn't follow this trend that's played is Snakewater, where the point is on the floor and there's height advantage directly above it. The lower part of snake's mid is actually really enclosed and there's nowhere to push all the way around the side of the point so teams are forced to push directly forward. In addition to this, the point itself isn't so easy to spam from the height advantage as it's actually directly above it.
I think if people have how they want the mid (and other parts of the map) to be played in mind when they design it, it makes it easier to come up with something people will like.
In the world of non 5cp 6es, I think people have less of an idea about what makes koth maps good as we only have one we really play. However, as koth maps are small and arguably the easiest type of maps to make, I'd imagine TF2maps.net has a wealth of koth maps nobody's ever even tested before because they weren't even designed with 6es in mind. Unless I'm mistaken, nor was viaduct though so it might be worth having a look through some of them just because there might be something that does it right accidentally.
I think ctf is even less well understood, but there have been some pretty radical modifications to the ctf format that could be interesting, I remember there was a map called fusion that allowed you to cap by bringing the enemy's briefcase to your own dropped one which could be interesting, although I'd imagine it might just discourage pushing entirely.