I suppose the first question is what state are the Indexes in, do you run regular Db maintentance jobs on the indexes?

If they are highly fragmented, then you could get very poor query result time, check them with DBCC SHOWCONTIG statement for each table/index if theres a low Scan density then a reindex might help increase performance.

The RID lookup to GMFPick isnt helping either so you might want to look at including the EmployeeIdNo,ProductIdNo, HouseIdNo, CustomerIdNo on the Nonclustered index (7 in your list), on the GMFPick Table.