Use git lfs for binary objects

[NathanJAdams]

An initial clone takes a long time and downloads 226MB of objects. Can only assume this is due to git lfs not being used and lots of deleted/changed binary objects in the commit history. An initial git clone produced the following:

$ git clone https://github.com/nrwl/nx-examples
Cloning into 'nx-examples'...
remote: Enumerating objects: 4273, done.
remote: Counting objects: 100% (1205/1205), done.
remote: Compressing objects: 100% (662/662), done.
remote: Total 4273 (delta 538), reused 1092 (delta 489), pack-reused 3068
Receiving objects: 100% (4273/4273), 226.83 MiB | 2.46 MiB/s, done.
Resolving deltas: 100% (2251/2251), done.

This can be fixed for future use by using git lfs for all binary file types

That won't remove existing binary objects already in the commit history though
To do that would involve either:

1. Editing the commit history
2. Creating a branch off the latest commit that didn't have any binary objects, copying everything over to it, using that as the new main branch and deleting any other branches that include the non-lfs binary objects