Consider adding this citation to reference/bibliography.rst and referencing it here.

The discussion of identifiability appears overly binary. A near-singular or approximately zero Fisher Information Matrix is not the only scenario implying practical non-identifiability. Even when the FIM non-singular, a very large condition number may indicate that certain parameter directions are only weakly informed by the data, corresponding to extremely small eigenvalues and unstable parameter estimates (practical identifiability issues). In such cases, uncertainty remains high and estimates may be highly sensitive to noise, limiting the extent to which experimental design alone can improve parameter estimation. Consider distinguishing more clearly between structural identifiability and practical identifiability arising from ill-conditioning of the FIM.

@snarasi2 This PR is only meant to organize the order of the documentation pages into the format outlined in the summary. The contents in guide.rst and overview.rst (for DoE) are what currently exist in the documentation page. This PR does not make any changes to the contents but establishes a general structure for the ParmEst and Pyomo.DoE documentation webpages, which will be later updated by @sscini, @smondal13, and @snarasi2. I will save these comments for @smondal13 and @snarasi2 to work on them during their future PRs to update the contents of the Pyomo.DoE documentation webpage.