Fix lesson for FilterInTrees

[apuignav]

After the last modifications, the lesson (https://lhcb.github.io/starterkit-lessons/second-analysis-steps/filter-in-trees.html) doesn't make sense as we're not really selecting those events that are not compatible with a K*. In the end, it's all very confusing.

What I did was reuse the decay to build D*+ -> D0 (->K*(->K+pi-)pi0) pi+ reusing the K+ from the original decay, but this is not very interesting. We need to think of a better physics case.

Ideas?

[dcervenkov]

@apuignav, glad you opened this issue. I was indeed baffled about where are the K* events being excluded. I was looking up the documentation to understand it when I found this issue. Even a temporary solution like linking this issue from the text would be great.