I can't say that I'm in favour of SGOTM teams starting from different positions, with different Traits, or with different free techs/improvements/what-have-you. Part of the fun of the competition is that the games are somewhat comparable. Having a team get a strong advantage before the game even begins (well, in addition to stacking some teams with more experienced players) due to a couple of comparatively poorer choices by other teams doesn't really sound all that fun to me.
There is an ongoing (almost finished) BOTM game where players had to choose their starting area. I don't want to discuss details of that game, since it is still ongoing, but I think that it is okay to comment (since one can make this comment from just looking at the initial screenshot) that players who picked different starting areas cannot really compare their games to each other. While it's a neat concept for an XOTM game (even though it comes at the expense of fracturing the already hard-to-compare games due to the lack of participation meaning that you and perhaps one other player may be competing for a given Victory Condition and in this case you might not even be able to compare your game to that other person if they picked a different choice at the start), I don't think that it's a concept that belongs in an SGOTM game.
Things that I did like from recent SGOTM games:
- Spammed Horse Resources in SGOTM 20, so that no matter where you settled, you got Horses, but so did the AIs, and they didn't make for great trading bait since the AIs could all easily get said Resource themselves
- Limited nearby Strategic Resources in SGOTM 21, with no Copper; Horse and Iron available even if a team delayed settling them but not within easy reach; and the 3 closest AIs not having a duplicate Strategic Resource to be able to trade one to us
- Improvements that exist on the map, but for all teams, such as a Farm from a "destroyed City" in SGOTM 20
- Settling locations for Horse, Iron, Stone, and Marble were all accessible for a relatively long time in SGOTM 21, so that most teams would have an equal chance to settle by all of them
- AIs starting with bonuses (Shaka getting an extra Catapult, an AI who isn't close to us getting a Worker at the start)
Those first two concepts were drastically different from each other, but they were well planned within the context of working well as a combination with the bigger picture of the given map. I.e. Horses being plentiful for the human player and the AIs meant that the Horses couldn't be abused for Resource Trades. Meanwhile, a lack of Resources for the human player couldn't be made up for by getting a good trade from a nearby AI, while each of the nearby AIs did have enough Strategic Resources to have the ability to build more than just Archers. But, for example, if Shaka had 2 Copper Resources easily available to him, but in some games he settled by a second one and in other games he didn't settle by a second one, that fact could have greatly altered the game due to some teams able to get Phalanx via good trading due to a random decision made by an AI--the map-maker did a good job in preventing such a situation from possibly occurring.
Thus, a game design needs to think of the big picture and whatever elements are incorporated should work together smoothly.
As for improvements existing on the map at the start of the game, make them a standard feature of the map for all players to experience.
As for simpler or not, which Difficulty Level to pick, etc, I don't think that these decisions should be made independently. A well-crafted scenario always takes the complete picture into account when each individual decision is made. And, yes, there can be several HUNDRED small decisions that can go into a map's design, so asking for feedback on one or two points would, in my opinion, only help you to decide between offering Complete Scenario A vs Complete Scenario B and thus such input should not necessarily cause a change in a given scenario unless that change is considered as part of the bigger picture of all of the other decisions that went into a map's design.