In this specific instance, the Channel Splitter
and Channel Combiner
are designed as a method for interacting with the red, green, and blue channels of a color image separately, not necessarily to be used on height data. That’ll be the reason you’re only able to process 3 at a time (and also, specifically 3, if any of the channels are missing inputs, the Channel Combiner
will fail to build).
Unfortunately, I don’t think there is currently a method of adapting that workflow to work in anything other that groups of three, just due to the nature of the data type it is expecting. Nor do I think there would be any practical method of implementing something like that, without defining some new image type with a dynamic number of channels.