The current review system is a bit unfair in the game. Whenever you try to get all 10’s in a row with the current system, you get three 10’s and one 9 due to one of the reviewers being mean to you. There is another situation where your trying to do a franchise of games like Madden, and you added new features in the new version of the game that weren’t added in the original version of the game that your doing the sequel for, the reviewers give you less than a 10 due to the game being the same, so the reviewers don’t care about the new features you added.
The 9.75 cap is there to reflect the difficulty of getting a perfect 10. Your 9.75 will have essentially the same sales as a 10.0 game would. That last .25 holdout needs to have something unique in your game… the kind of something only dedicated specialists can provide.
Also look at it in real life. Take Will Wright, of Maxis fame. He makes the original Sim City. It ends up being a smash hit and lets his Company Maxis take off and grow. Was Sim City 10.0 perfection? Nope. I’m a massive Will Wright/Sim City fanboy and I’m still willing to say that the original Sim City was not 10.0 perfection. That didn’t stop it from setting an industry standard and defining a new genre.
Will didn’t have the raw game-making experiance to really hone his game to absolute polished perfection.
As for the making a sequel to a game with features turned on but the same engine… that is more a memory issue. It would require the game to keep track of exactly what features were used in every single game and compare them to every single feature used in a potential sequel. Far easier to say “is the sequel’s engine newer?”
Besides, the reviewers might be asking “If your engine was capable of [Feature] back then, why didn’t you use it with [Original game]?”