I would welcome opinions on whether FTS would help in our scenario.

The table has about 28 million rows and has about 10,000 rows added every
morning in an intensive batch process. Then another batch query process is
run that looks for partial name matches.

Eight of the fields we query against are char or varchar, all with single
words.

Sample query: SELECT casenumber FROM table WHERE fname LIKE 'j%' AND lname
LIKE 'smi%'.  Rreturns 55,874 rows in 7 seconds in QA.

I am also concerned about time to populate FT index every day.

Machine is dual Xeon 3.2 with 4GB ram, OS on RAID1 & SQL data on RAID5. Win
2003 Server and SQL2000, with SQL2005 later this year.

Thanks much!