Add missing subgroup formative element descriptions

[brownag]

This is a first cut at addressing the unexplained taxa from #5 by identifiying shared criteria from taxa with same subgroup-level elements.

If an element is used more than once I tried to summarize the shared criteria ignoring specifics that are only for some taxa. If a term is only used once and it refers to another taxon name (e.g. "aquodic") the description is simply "like aquods" or equivalent for now.

Often with the remaining missing subgroup-level modifiers the definitions are not consistent across all Great Groups where they are used. There is a need to strike a balance of generalization to make reasonable explanations. Formative elements at subgroup level often either re-state or rely on information from higher taxa criteria or do not really pertain to the criteria used at subgroup level (at least in name), so sometimes it produces unsatisfactory explanations (from the perspective of formative element meanings alone)

For instance:

• We have many different varieties of Durids such as the "Xereptic Haplodurids" whose subgroup modifiers ("Xereptic") require less than very strongly cemented duripan. "Xereptic" does not say anything in formative elements about duripans but is modifying the "Dur-" used at suborder level, with the 'Inceptic' aspect suggesting "limited development" of Haplodurids vis a vis cementation class.

• in the "Argiaquic Argialbolls" the "argiaquic" modifier means that the soil does not have an ATC between eluvial horizon and an argillic. The formative elements "argi-" and "aqu-" (which are both implied in "Albolls") give no specific information about the requirement for ATC in Typic subgroup (implied by filtering out non-ATC soils before "Typic"). It is cases like this that prevent the use of higher order formative elements to directly determine meaning at subgroup level. The formative elements take on special meaning depending on the Great Group they are combined with and the taxon that they are grading to.

[brownag]

Closes #5, discussion on #17 for any future plans for updating any of the formative element dictionaries (which now have their missing values filled in)