Update externals based on ESCOMP/CESM tag 'cesm2_3_beta17'

[gdicker1]

This is the first of a series of PRs that will update EarthWorks to more closely match the current development of ESCOMP/CESM.

This PR combines changes to use externals based on those used in the cesm2_3_beta17 tag and to update the cime_config folder with CESM changes.

[gdicker1]

Sub PRs:

• https://github.com/EarthWorksOrg/ccs_config_cesm/pull/25
• https://github.com/EarthWorksOrg/cime/pull/11
• https://github.com/EarthWorksOrg/CAM/pull/28
• https://github.com/EarthWorksOrg/CMEPS/pull/9
• https://github.com/EarthWorksOrg/CTSM/pull/7
• https://github.com/EarthWorksOrg/CTSM/pull/10

[gdicker1]

Note this comment above was updated to reflect the changes in 7563b99

(i.e. that the EWOrg/CTSM PR has been replaced)

[gdicker1]

After making the updates (to EWOrg/CTSM), I think this PR is ready to be merged.

Testing the ew-pr category against runs of the same test category using the ew2.3.001 tag revealed:

• QPC6 with nvhpc had NLCOMP and DIFF fails [^1]
• CHAOS2000dev with nvhpc, gnu, and intel had NLCOMP and DIFF fails
• F2000climoEW with nvhpc had NLCOMP and DIFF fails
• FullyCoupledEW with nvhpc had NLCOMP and DIFF fails
• FHS94 with nvhpc only had NLCOMP fail
• FKESSLER with nvhpc only had NLCOMP fail

I think in light of CLUBB namelist changes and development in CTSM (namelist and otherwise) in cesm2_3_beta17, these failures make sense.

[^1]: NLCOMP fail means some namelist(s) changed between the test and the baseline. DIFF (also BASELINE in output) means that differences appeared when comparing output from the test and baseline.