Everything else being equal, the response times for querying these two tables individually should be nearly identical. The difference would only be whatever time it takes to package the additional 50 fields for routing. That will take some time, probably uncalculable per row; But if you are returning 1,000,000 or so rows, that might be significant. For similar queries on these two tables, the execution plans should be indentical. If the queries are different, or if one table has indexes the other does not, all bets are off.
Now if you need to join these two tables, you could create a CLUSTER based on the 2 common columns. That way, the data from the two tables would be physically stored together in the database files, which would drastically reduce the physical read times for such queries. This adds a small overhead for DML on these tables, but this is much more than made up for on queries.