improve table output for lavResiduals

[vsavalei]

Feature request: lavResiduals(fit,output="table") does not include mean structure residuals, and does not respond to options such as zstat=TRUE (another column could be added to help if top residuals have large z-values).

[vsavalei]

More (should I just move all of this to #238?):

• I wonder if the components of the $summary output from lavResiduals should be (also/instead?) computed by fitMeasures. The separation of the SRMR into cov and mean (or threshold) structure in particular is very helpful but cannot be requested from fitMeasures and is harder to grab from here.
• For multiple groups, the summary output from lavResiduals is by group, but not overall. Overall would also be helpful?

$`Grant-White`$summary
                          cov   mean total
srmr                    0.077  0.087 0.079
srmr.se                 0.008  0.007 0.007
srmr.exactfit.z         4.263 10.127 5.436
srmr.exactfit.pvalue    0.000  0.000 0.000
usrmr                   0.063  0.085 0.067
usrmr.se                0.013  0.009 0.011
usrmr.ci.lower          0.041  0.070 0.049
usrmr.ci.upper          0.085  0.099 0.085
usrmr.closefit.h0.value 0.050  0.050 0.050
usrmr.closefit.z        0.969  3.899 1.530
usrmr.closefit.pvalue   0.166  0.000 0.063

[yrosseel]

All good suggestions. A rewrite of lavResiduals.R is long overdue. I will bump it up on my TODO list.