I also remember blaze directly eval'ing Python pre-Skylark, but I think the bigger problem was build hermeticity. This was compounded by env-related issues making it hard to build Python hermetically in the first place, and the importance of cached object files and build outputs in managing the incredibly large monorepo that is Google's source tree. And that's all before considering the Py2->Py3 migration (which, fortunately, was tackled afterwards).

I think a new bespoke DSL would have been a non-starter since so many build scripts had already been written by the time Skylark was being conceived.