Reduce rerun_c library size (by depending on fewer unnecessary crates)

[Wumpf]

What

• Fix Make rerun_c smaller #2905
  • reduces Rerun C size from 41.0MiB to 34.2MiB (MacOS aarch64)
  • feels like there's more we should be able to do, but this PR stops the bleed of having it become larger as we add new types!
  • note that since this is a static library, typically the actual stripping occurs when linking the final binary, so while it's great that our library gets smaller for many reasons, it's not critical
• re_log_types no longer depends on re_types
• re_sdk no longer depends on re_memory
• re_sdk no longer depends on re_types
  • things only come together now on rerun crate

Dependency graph before:

Dependency graph now:

Graphs via: cargo depgraph --all-deps --workspace-only --all-features --dedup-transitive-deps | dot -Tpng > deps.png
Rerun C checks via: cargo build -p rerun_c --release && ls -l target/release/librerun_c.a

Checklist

• I have read and agree to Contributor Guide and the Code of Conduct
• I've included a screenshot or gif (if applicable)
• I have tested demo.rerun.io (if applicable)
• The PR title and labels are set such as to maximize their usefulness for the next release's CHANGELOG

• PR Build Summary
• Docs preview
• Examples preview
• Recent benchmark results
• Wasm size tracking

[jleibs]

Nice. This seems logical to me and paves an improved path for standalone packages for adding new types without cluttering the sdk/rerun_c.