But.... A clustered index on a primary key will usually return a single record in two or three page reads. The same query will do three or four page reads if a non- clustered index is used. However, a query that return 100 records will typically take four or five page reads using a clustered index. The same 100 rows would typically take 95 or so page reads if using a non-clustered index.
A clustered index makes a minor improvement to a point query, but a major improvement to a range query. So, I don't claim that a clustered index won't help a single-row lookup. But, I do claim that your one clustered index should be used where it will make a huge improvement, rather than where it will make a small improvement. Even if 95% of your queries are primary key lookups, it is still a bad idea to spend your clustered index there.