Hey there! First let me know if this is not the right place for discussions like this, I'm happy to move it elsewhere if so :)

Recently I've been working with a large Tree and with treeData I can think about two main ways of structuring it: 1. A memoized object derived from your tree dependencies; or 2. A Tree state / reducer with the triggers to update it (main case in docs). The problem I see/I'm facing with it is that it translates into a linear structure (coming from React / how hooks work), being transformed into a Tree:

function Component() {
  const firstLevelData = useGetFirstLevelData(); // even more of a problem when these are async
  const secondLevelData = useGetSecondLevelData();
  // + any other dep to build the tree
  const treeData = /* build tree, with everything needed inside this component */;
  return <Tree treeData={treeData} />;
}

This got me thinking: the DOM itself is a Tree, so having it performant and with good ways to abstract it is part of the problem React had to solve, which makes React nodes potentially the best way of solving this:

function SecondLevelNode() {
  const thirdLevelData = useGetThirdLevelData();
  return <TreeNode />; // + any other Node component, can even be <LoadingNode /> when async for example
}

function FirstLevelNode() {
  const secondLevelData = useGetSecondLevelData();
  return <TreeNode>{secondLevelData.map(() => <SecondLevelNode />)}</TreeNode>
}

function Component() {
  const firstLevelData = useGetFirstLevelData();
  return <Tree>{firstLevelData.map(() => <FirstLevelNode />)}</Tree>
}

I know there's a deprecation note in

1. Is there any limitation/issue for it to be deprecated?
2. Does it make sense to keep supporting it for better structuring Trees with different dependencies for each level?

Thanks!!