@tom T: Your list is a classic 4cmc gas-list. Nothing wrong with it. And looks solid. To answer your question:

1. On sites like Deckcheck.org there are Lanstill lists similar to the one below, but this thread seems to vary a lot with lists. This is quite confusing: why would you all go in different directions while the 'top8-lists' from deckcheck all look almost the same?

Because Landstill is a metagame control deck. One does not pilot the 'best' Landstill list because there is non. In fact, a 'standard' Landstill list is either:
1) Have good matchups against decks x,y while losing horribly to other decks (e.g. classic Landstill beats aggro then dies to burn/combo)
2) Have good matchups across everything so becoming very tough to pilot correctly (e.g. IMO Countertop Landstill/Hanni's List, Chii's Scepterstill list). These are the non-normal non-classic lists that have quite a different selection of card choices.

Inherently, all these forms and choices are decisions made in response to the metagame. If there's more tribal around, then UWr Landstill is the better deck with REBs against Merfolks, Firespout killing a ton of crap. UWb is solid against slower Countertop meta with Bantish decks since UWr does nothing here (red does nothing big against these decks). If there's a ton of combo/burn, then the classic list just fails hard i.e. I remember trying so hard to beat burn, and I came to the conclusion it was about countering all burn, and drawing Pulse of the Fields. But instead of playing a classic list, if you just tweak some cards to incorporate Countertop it becomes much easier.

Now, if you disagree with my statement, just take a look at the very nature of Landstill. It has no fixed win-condition/gameplan. We say we win with Jace 2. But any deck can win with Jace 2. We beat with Factories, but so does Dreadstill. The deal with Landstill is: There's no fixed gameplan, it's inherently a control deck, and a control deck only does its best when metagamed. Unlike Dreadstill or Countertop, these decks actually have a gameplan e.g. winning with Dreadnought, or locking with Countertop. That's the reason why there's no universal best landstill list. In the same light, there's no universal best Dreadstill list, or Zoo list. It's all tweaked, but Landstill has a much higher degree of freedom when it comes to deck design, which explains the huge variants of cards played in various top8s, anything from Fact or Fiction, Decree, Crucible, Elspeth, Ajani, Jace 2, Eternal Dragons, Counterbalance, Scepter, etc



Furthermore, the primer doesn't have a matchup-analysis.
2. I wondered which decks UW(x) Landstill was good against?
Classic Landstill (e.g. your list): Aggro, Goblins, Bant, Countertop, Stax/Stompy (we have higher basic count and more lands compared against other control decks that rely on less basic/lands/more lower-cmc spells)

50/50 matchups: Zoo, Enchantress (this one is tough), combo
<50/50 matchups: Burn, Aggro Loam (this could be 50/50, depends if they draw the nuts)


3. What strategy should the Lanstill player pull of against Zoo, Goblins, Merfolk, AggroControl (as in New Horizons and Bant Control)

In that meta, UWr list would do well. Firespout kills Gobs/Merfolks. Run no less than 8 removal spell + 2-3 EEs, and things should be easy from there. Landstill inherently beats Aggro-control, that's its strength. However, against such a meta, I would not drop below 4 basics because all of these decks run either Wasteland or Price of Progress. Against Zoo/Gobs/Merfolk keep the board under control, EE the Vials, and stick a Planeswalker, against Bantish decks, play around Daze/Stifle, that way you don't lose tempo and you out-tempo them when they waste their turns holding back creatures. When they finally drop creatures, you have mana open to initiate counterspell/removal and you should win the counterwar since they have tapped out to play threats so you have more mana advantage. If that fails, you untap and play another removal spell or a Planeswalker that buys cards/time over turns.