That new team has also got a salary 57xx player who is a 'hall of famer' - a bit old though (26), granted.
In this age the potential is a joke, and even if he got 5700 salary and got some respectable skills, you wouldn't train him in the areas where potential makes a different.
The second new team enters at rank 8! It also has a worse roster I think, and the best player is somewhat worse, at 'MVP' (also a bit old).
The potential ain't the player abiilitys, a bench warmer could be the MVP of the league - only the names for then are a bit confusing they should be more related with their potential not with their abilities(you could draft hall of fames, who can't play).
1) an inactive new team should after 3-4 weeks be significantly weaker than an active new team. Maybe remove their game shape training. I don't object to bots being competitive mind you.
You could also train game shape, and you got an little average ... But if the time difference is so big, that you got a big advance after four games you got big problems to catchup older teams, which should be posiibel.
2) new teams should join their league tables consistently from the top to the bottom rank - so the lucky starter mentioned above would still promote, good luck to him.
In a league where no teams play this games for more then a month, you need luck to promote ...But i don't think that you earn that luck more then the other player, just because you play one month longer.