The problem is fundamentally that doing a thorough review of every app is unreasonably expensive and as a consequence, that will not happen.

You are then left with a trade off between false positives and false negatives. It is entirely possible to have both at once.

The real problem is that centralizing this process is in error. Because not all users are the same.

It makes sense to have a manual review process for e.g. games, where anybody can make a game but games are relatively expensive apps that can eat the cost of a manual review in exchange for the largely uninformed user being assured that the code is trustworthy. The review process could then e.g. exclude apps that screw with the operating system in order to implement some kind of shady DRM.

Some users are using their device in a professional context where lives are at stake if the device gets messed up and can justify an even more thorough review of anything that goes on the device.

Some apps don't require any meaningful permissions and if run in an appropriate sandbox should not require any review at all.

Some users are technically inclined and would like to run whatever code they want on their device, including code that modifies the operating system and code they wrote themselves.

There is no one size fits all. A system that pretends otherwise satisfies no one.