Well the key is via the use of components using the context based animation concepts rather than the migration pathway used in animation blueprints. The use of components fits with the new game play and features plugin as the plugin handles the issues of dependencies as to the fundamental requirements of migration path driven animation so it can be assumed that the requirements of animation state changes could be don using these two new framework building tools.
As to applied functionality proof of concept still needs to be provided as to the replacement of the animation blueprint that up to 4.27 has been more of an extension as to the codeing requirements rather than doing it’s job of simply changing the current state, with out evaluation, based on an event that has already occurred based on the data supplied but what ever controller is being used.
Reinventing the wheel though will take some time as in most cases Epic by providing simple examples is stating that this is the way things are done rather the stating that it just a base in which to develop more complex design. Example of this trend is the use of “in air” for jumps when in fact the state changes should be based referencing the ground, IE as in being grounded 
With out writing a book Epic has stated that the current design logic is flawed, as I pointed out, in what should be considered a next gen video game engine yet sill makes use of technology logic that has been with us since the 1990’s.
This is where the need for components as a staring point comes is as it’s a fundamental building block as to the “requirements” of a top down modular design comes in that excludes the need for dependencies, such as animation blueprints, in favour of data friendly and tree based logic of just making it work.
As an example of concept.
This
Follows the logic of top down as it is a component that is added to the controller and not the animation BP in the traditional usage of migration pathway driven animation.
It is by design context driven of the states necessary to climb a ladder
It is added directly to the controller so no need to cast or play the 20 question Boolean game
It does not require the use of state machines as it gets it’s current state directly from the game controller.
It’s dependencies is limited so easy to adjust to as to custom requirements.
It’s data driven
It functions as part of a logic tree
Best of all components scale like crazy as the requirements is in the component with out the need to figure out how to wedge the expansion into the migration path.
LOL hay sorry but you asked but as a TA there are ways things should work as to usability and although animation BPs in 4.0 was OK in 4.27 they should have be replaced long ago using Star Trek technology.
To restate my opinion Unreal 4 is not broke as much as it’s work in progress so by knowing that it’s up to the end use to decided if they wish to go for the ride, in Epic’s defence has also stated as such, but here is hoping that UE5 is the advancement phase of achieving “Next Gen” status as far as animation subsystems goes.