EVI recommendation

I am working on top Index Advisor recommendations and
have a recommend to build an Encoded Vector Index on
a field which is declared as the single unique key to a file
used daily for several applications.

This file was created some years ago using DDS and has
about 1.3 million rows.

My questions:

1: To meet the Index Advisor recommendation, would it be a
good idea to create an SQL index on the recommended
field/column, or better to re-create the DDS file with DDL ?

Lazy programmer wants to just add an index, but diligent DB
admin thinks "pure" solutions give better performance ?

2: An MC Press article by Kevin Forsythe January 25, 2004 seems
to say EVI's are better "where there are few distinct values",
but this file has over a million distinct values, so lazy programmer
thinks this is a deal-breaker ?

Lazy programmer thinks this may be supported by October 31,
2013 developerWorks article: "Accelerated analytics - faster
aggregations using the IBM DB2 for i encoded vector index (EVI)
technology, but lazy programmer also thinks Index Advisor is
his friend ?
Note: the second paragraph of the Overview seems to support
Kevin Forsythe's advice.