Ray -

Since there is no clustered index on this table, I would assume that data rows are stored sequntially.  Regardless of what is in my select statement, as long as I have the right where clauses, the query engine should use index seek to get to the addresses of the desired rows and then get the information of all the columns for the select statement from the actual data page address location.  FYI, the where clauses in my case return about 7.8% of data rows.

What do you mean by when you say that the optimizer will have to make an additional trip to get the other data requested in the select statement?