The position transformation of a model produces a incorrect vertice position -.- For example one vertice of the model have a y-position of 0.125. I add an duplicate of this model and transform the model y-position with +30. So the example vertice should have y-position 30.125 but it doesn't. It has position 30.12499. Whats going wrong with the programmers of blender that a simple addition doesn't work...
I suggest you learn a bit about mathematics in computers, read e.g. wikipedia page on floating numbers… To summarize, computers have limited precision handling of numbers, so (especially with floating point ones) all operations will add some small error - and 3D stuff implies a lot of operations. No bug here.
We choose to live with this limited implementation, even though it has the consequences as you witness.
The alternative is using double precision floats, which will nearly double the amount of memory needed for storing 3D data in Blender.