1) The answer to this for any language depends upon whether or not your approach is descriptive, in the way that many a reference grammar is, or "linguistic" (i.e,. consistent with some theory of language such as Head-Driven Phrase Structure Grammar, Construction Grammar (proper), Cognitve Grammar, Government & Binding, X-bar, Minimalism, Word Grammar, etc.; more generally I mean a framework like cognitive linguistics, functional linguistics, linguistic typology, generative linguistics, under which specific theories of grammar are subsumed).
2) Some grammarians (notably Bill Croft) have advocated that syntactic categories (nouns, verbs, etc.) are language specific and as he adopts construction grammar as a theory of grammar (Radical Construction Grammar), grammar is internal to constructions. Hence Joachim's advice is not just useful but supported by linguistic theories such as construction grammars, HPSG, (New) Word Grammar, etc. (see e.g., http://www.cfilt.iitb.ac.in/gwc2010/pdfs/15_tamil_wordnet__rajendran.pdf & http://www.sas.upenn.edu/~haroldfs/public/h_sch_9a.pdf). Grammars like Schiffman's A Reference Grammar of Spoken Tamil or Lehmann's A Grammar of Modern Tamil use antiquated terminology adopted from Latin grammarians (see the second link above for the problems with attempts to apply the "classic" cases to Tamil). Same with verbs (see e.g., http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2928482/).
3) For some discussion of Tamil syntax among many other aspects of the Tamil language. see here: http://www.infitt.org/conference_papers/ti2009_conf_papers.pdf#page=275
1) The answer to this for any language depends upon whether or not your approach is descriptive, in the way that many a reference grammar is, or "linguistic" (i.e,. consistent with some theory of language such as Head-Driven Phrase Structure Grammar, Construction Grammar (proper), Cognitve Grammar, Government & Binding, X-bar, Minimalism, Word Grammar, etc.; more generally I mean a framework like cognitive linguistics, functional linguistics, linguistic typology, generative linguistics, under which specific theories of grammar are subsumed).
2) Some grammarians (notably Bill Croft) have advocated that syntactic categories (nouns, verbs, etc.) are language specific and as he adopts construction grammar as a theory of grammar (Radical Construction Grammar), grammar is internal to constructions. Hence Joachim's advice is not just useful but supported by linguistic theories such as construction grammars, HPSG, (New) Word Grammar, etc. (see e.g., http://www.cfilt.iitb.ac.in/gwc2010/pdfs/15_tamil_wordnet__rajendran.pdf & http://www.sas.upenn.edu/~haroldfs/public/h_sch_9a.pdf). Grammars like Schiffman's A Reference Grammar of Spoken Tamil or Lehmann's A Grammar of Modern Tamil use antiquated terminology adopted from Latin grammarians (see the second link above for the problems with attempts to apply the "classic" cases to Tamil). Same with verbs (see e.g., http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2928482/).
3) For some discussion of Tamil syntax among many other aspects of the Tamil language. see here: http://www.infitt.org/conference_papers/ti2009_conf_papers.pdf#page=275
I don't get your question. There is no need to develop grammar rules (?) for such an ancient language. Rules (i.e. illakanam) tends to exist thousands of years before. May be your questions should be on either understanding grammar or interpreting grammar of Tamil.