G'root, which I understands as generalised Taproot, was announced a year and a half ago. I'm wondering what the development is on it and if it is part of the taproot softfork.
It is not part of the BIP drafts we're working on for Taproot, and there has not been any development on it apart from publishing the basic idea.
The reason is simply that there are far too many ideas to include all at once; ultimately everything boils down to costs/benefits, and I think just G'root is only marginally useful. And softforks are more than a technical process: we don't need to just propose a technical improvement, but also find adoption for it. The more complex the ideas involved are, the harder the latter becomes.
As I explained on the ML a while ago, I think that G'root does have actual tangible benefits when combined with cross-input aggregation. We deliberately chose to not include that in the current taproot proposal as it interacts in complicated ways with
OP_SUCCESS, but perhaps a future change that includes both is possible.