Support missing weight data on consensus level.

[shamil-gadelshin]

Relates to fast-sync engine: #2754 (comment)

[dariolina]

Is this critical for fast sync to work or a nice to have?

[nazar-pc]

This allows us to not have a bunch of code written otherwise. I'm not sure why this issue exists though, I'd expect a PR right away instead, it would be basically one to three lines change in sc-consensus-subspace.

[shamil-gadelshin]

I'm yet to discover these lines of code. This is why was going to research it right after the current PRs are done.