Yes I completely agree and its very reasonable that defaulty when user wont choose any specifc document type, the engine shouldn't be overloaded by retrieveing all possible data. I am only arguing with the naming convenction :). I am just saying in case of usability, that filtering is not exactly equal to listing all needed conditions. Once learned you know how to use it, its just may make some additional confusion, when you use it for the first time.
So no offence, its good as it is, as long as it works :) Thanks for the reply!

To add to what Juraj has said, choosing the document types you want also gives you a way to make sure that the transformation you choose will work when it contains Evals for fields that may only exist in the content type you wish to display. In other words, you should make sure that your transformations match your document types.

On that note, if you select more than one document type, you would need to add some fancy code to make it change the transformation based upon the content type if you wish to use properties that are not common across all of the document types that you have selected. You would basically need to have a nested repeater and a function to evaluate the NodeClassName property of the document and translate that into a transformation name to feed the nested repeater. This may be getting ahead of you as you are new to Kentico, but it is possible that you will run across an scenario where you would like to do this in the future. There are other ways to accomplish this goal, but you can cross that bridge when you come to it.