Let us go back to RDBMS basics. A row models a complete fact ("John
bought a squid on 2007-05-12"). This fact should appear in one
table, one time and in one way in the entire schema. This is the
foundation of normalization.

In the old days, with punch cards, paper files, etc. we would
physically move these physical records from one physical location to
another physical location. We had redundancy and we want to get rid
of it. You want to increase it.

You should never use IDENTITY in an RDBMS; you want to have a
relational key. A key has nothing to do with the physical location of
the data in the hardware; it is based on the nature of the data being
modeled.

You entire approach is that of someone managing a 1950's paper file
system. Please read a book before you code again. If you want
stinking dirty kludges, then you can get them in Newsgroup; but being
a good RDBMS programmer will take YEARS of hard work. Be better than
that.