From my view small real workflows and tasks someone could follow might do the trick if people are interested in how things work and maybe reluctant to try out new things.
Concerning knime and excel - you could substitute some functions and make them more effective but they also could work together.
You are aware of the free book helping people to go from excel to knime
And then there is this meta workflow that list a lot of things how to handle excel related questions in and with knime.
Concerning databases and big data KNIME has a lot of great functions and connectors and can handle the steering of spark and pyspark functions (I plan on releasing another meta workflow about that soon).