January clean-up

[odunbar]

Purpose

Sort a variety of low-priority tasks that have been accruing

• Closes Docs still mention bors #433
• Closes Doubled warnings #409
• Closes Multiple ensemble members identical - request for better error message #421
• Closes Alternative approximation of std-approx in SECNice() #413
• Closes Remove the Unnecessary inflexibility in building the sampler #423
• Closes A few potential README improvements #408

Content

• Replaces mention of bors with squash and merge request
• elseif in warning about N_ens/N_par to remove duplication
• added column check for G_ens in update_ensemble, if non-NaN columns are overlapping, a warning is produced indicating this prior to LinearAlgebra issues. Unit tested as well.
• improves readout for FailureHandler and Localizer when verbose is used (before this did not show the actual method used)
• new SECNice approximation is accurate to 10^-4 on average to the prior approximation and requires no precompute!
• Updated Localizers example to new defaults and for readability
• get_obs obtains the length of a vector by summing piece lengths, not estimating it by multiplying the first vector length
• Added copy-pasteable(!) example into the Readme complete with forward map data etc. for users to play with.
• Moved citation down and added a link

Validation

• Unit tests for the G_ens multiple columns the same
• Verified SECNice accuracy with localizer example and direct approximation comparison
• Readme can be viewed here
• All checks/tests pass

---

• I have read and checked the items on the review checklist.

[odunbar]

As I have condensed a few small changes in here - I will merge when all tests, coverage and other checks pass.
I will reach out specifically for the README review after the merge has occured so we can build and improve on this initial attempt in a future PR and it does not hold up other bug-fixes.