I have found this in a university text book and have been told it has many erros. What is wrong here?

I have found this in a university text book and have been told it has many erros. What is wrong here?

The $Pb$ and $Qb$ magically appear and $\forall x(Px\to Qx)$ is improperly instantiated. The $(\neg Qb\wedge\forall x(Px\to Qx))$ should be negated (as should $Px$) and it's treated like $\vee$ instead of $\wedge$.
Each path of a tableau kind of represents a different scenario, so that what happens in one path doesn't influence what happens in another beyond the point where they branch.
Here's what I get:
.
This tree is fine, but it's not a tableau, it's syntax tree. Whoever told you it's wrong is confusing tableaux (which represent ways a formula can be falsified) and syntax trees (which represent the syntactic structure of a formula, breaking it down in simpler sub-formula until you reach atomic formula).