The #1869 PR introduced support for new filters and nested properties together.
While that is a valuable feature, it might be risky for some runtimes, introduces challenges for some tools and other runtime implementations, and could be replaced b powerful alternatives.
This issue aims to assess support for that feature, without necessarily following the approach adopted by #1869, and to consider a long-term path that preserves Liquid’s stability.