[iabacu]

If the over-engineering is in the wrong dimension (one that the project doesn't need), then the cost will be double: adding the over-engineering and then detangling from it.

The problem is there are too many dimensions a given project can be over-engineered if the future is uncertain. So even an educated guess has a good chance of being wrong.

[iabacu]

Even if the guess is reasonable and is eventually true, like "we'll need to scale, so might as well prepare for 100x capacity now", it's still frequently a mistake to over engineer.

The architecture to support something that is not needed tends to introduce rigidity into the codebase, adding a tax to future changes in order to maintain those features.