Even if he wasn't correct, trying to show bias of some sort is the least promising alternative we have. Why?
1. In some proof scenarios it would require you to get some official to admit his/her bias or to somehow climb into his/her mind to find the bias yourself ...
Not likely. ;-)
2. If you go for statistical aggregates, it begs the opposition to insist (correctly) that you must eliminate all other possible ways to account for the aggregate differences before you can win the point on the basis of there being no explanation left but bias ...
Not likely.
We should treat bias (which probably actually is present ... it's just not provable) as a simple statistical artifact of a system that presumes inequality (i.e., under the current system SOMEONE is going to lose ... that someone is called the NCP ... and that someone loses 90% -- or whatever -- of the time).
The way to argue this isn't bias, it's fundamental rights that find specific protections in the Constitution (which is a body of common law in the UK scenario). There are two fundamental rights implicated:
1. The fundamental right to parent your children.
2. The fundamental right to a fair trial -- which none of us are getting under a Preponderance evidentiary standard. The implicated right requires a Clear & Convincing evidentiary standard. This means that there's not only a Procedural Due Process hole in the proceedings, the trial itself is fundamentally unfair and deprives you of your fundamental right to fundamental fairness.