- Consider use of SQL::Splitstatement in Deploy

- Convert all exported subs into non-anonymous package subs

- check for nothing returned from select statements: what should be
  returned by hashes, arrays, etc?

- implelent qi() for quote_identifier.

- Use the {Statement} attribute for recent SQL sentence

- Handle this properly:

    where => [
        { status => [ 'Active', 'Expansion', 'Greenfield' ] },
        ' OR ', { status => undef }
      ],

    WHERE
        status IN (?,?,?) OR status IS NULL

- drop_tables() should be done with "DROP TABLE" statements

- Create a FAQ section based on the questions you (or the typical
  first-time viewer) might have about this module. Obviously you
  wouldn't have all the answers, but I would be happy to accept a pull
  request of questions and then fill it out myself as needed
  afterwards.
  
- Audit the documentation for missing, wrongly-named, or removed
  methods and functions. I know that all of the sql_* functions for
  example have had the "sql_" part removed. Once again, I wouldn't
  expect you to write the actual documentation unless you really have
  nothing else to do, but it would certainly help me to know what to
  edit next.

- Make the examples in the SYNOPSIS section more consistent, or with
  better variable names, or with an example of your own, or perhaps
  with examples borrowed/stolen from other DBIx modules, depending on
  your taste.

- Refactor _ejoin() and _query() functions to remove the duplication
  that exists between them. Not an easy task I will admit, and to make
  it worse I am also sorely missing some tests here to reliably detect
  input/output combinations.
  
- Perhaps run Devel::Cover on the tests and see what isn't covered. I
  know that the transaction stuff definately needs some work. [Not
  something for a pull-request I admit]