I don't necessarily agree with the post's use of NVARCHAR for all VARCHAR based data types. That would impose a massive performance decrease in general for all String columns. Users should opt-in for this use by either using @Nationalized or the config property hibernate.use_nationalized_character_data.

Other data types this blog post outlines were introduced as part of SQLServer2005Dialect. This may have been done due to an oversight on the original SQLServerDialect and rather than impose concerns with schema generation backward compatibility, we elected to maintain these as part of SQLServer2005Dialect going forward?