Various updates before submitting to CRAN, including those from issue #178

[svteichman]

Various updates before submitting to CRAN, including those from issue #178.

• updates the description of the return object when using ?emuFit
• adds a "quick start" vignette to give a quick example of running emuFit() and interpreting results
• adds X_cup_from_X_fast() and get_G_for_augmentations_fast(), two functions from PR Speed up augmentations #148 which is now pretty out of sync with the codebase
• slightly updates the information the user gets when verbose = TRUE
• skips the test from fit_null() because this is slow and deterministic, and fit_null() is tested extensively in "test-fit_null_symmetric()" as the comparison method

@adw96 right now the quick start vignette doesn't have an example for clustered data, mostly because we don't have real data with clustering and I wanted to make it as streamlined as possible. However, let me know if you want me to add one, and I can just generate fake clustering patterns for the wirbel data and fit a clustered model as an example.