In order to have a 5 card deck, native village is mostly a no go for insane level abuse. Sans outpost, you cannot play the cards on the mat. You might be able to come up with some sort of pull cards off, generate new copies, trash the originals, and put the new copies back on the mat, but the 5 card deck cap is a bit harsh.
Sans outpost, this means you need a NV at the beginning of the turn, and a NV at the end of the turn. That leaves 3 cards to trash and regenerate multiple cards - not so easy to do.
Using outpost/nv is cute, but still has practical problems. This means that outpost takes up one of your slots (leaving you with 4 slots). Unlike haven, KC/NV takes up two more slots. This means that during your outpost turn (having played your payout cards during the first turn), you have only 4 cards you can play. KC/KC/NV/NV means that at most you could have six cards that are actually played on the mat.
I think we can modify BB's solution into something like the following nightmare for hard mode (I don't have time to work through this for optimal setup/play at the moment):
Starting hand:KC/goons/watchtower/outpost/scyring pool
On the mat: KC/KC/KC/KC -> 9 havens -> holding the following 27 cards: 5 KC, 9 goons, 9 throne rooms, scheme, and 4 markets?
So my playable hand begins with playing out the 9 TR in an efficient tree. Off of that I will play a KC tree. At the end of the KC tree I will play a whatever combination lets me get down all 10 goons with as many buys as possible with the following two caveats:
1. I have to have an unplayed watchtower to trash my buys of mass copper.
2. I have to have a spare action to play scheme -> scrying pool at turn end (these can be played off KC or TR branches if needed).
I then buy to trash how ever many dozen coppers I'm allowed and play the outpost. I place Scrying pool on deck top and draw it to start the outpost turn.
My outpost turn consists of playing the pool to draw everything -> KCx4 -> haven to regenerate the deck and then KC/goons for 4 more points with watchtower trashing.
I'm not sure if it is optimal to use all the TR's, but at least a few should be good to KC more buys, so the optimal mix of cards may have more markets and fewer TR. Likewise, the optimal mix of KC and TR used during the normal turn vs used to bulk up card storage (and hence buys)I leave as an exercise for the reader.
I haven't calced it, but I believe this should break 150 VP chips/turn without too much trouble.