Two of my most recent reads were A History of Modern Greece by C.M. Woodhouse and Alexander the Great by W.W. Tarn. Each has value; Woodhouse endeavours to document Greece from 306 C.E. up through 1977, while using just 300 page; Tarn almost single-handedly launched the modern study of Alexander III. They also have deficiencies. Most glaring for each is their tendency towards racial, ethnic or other stereotypes, combined with unabashed judgments about people or events. Tarn believes that Alexander was the first person to suggest the universal brotherhood of mankind, began the two main trends in political theory (monarchism and universal brotherhood, from what I can tell) and that if Alexander hadn’t died, we would be living in a Utopia under his descendants; Woodhouse hates Mussolini with a passion and is perfectly willing to deride him as an absolute monster without justification other than that he attacked Greece and felt petty jealousy toward Hitler.
If I actually had to use these two authors as sources, I would find myself in a deep dilemma. On one hand they know what they are talking about for the actual narrative, on the other, their arguments and their methodology is weak (in fact, I would go so far as to say that Thucydides, writing thousands of years earlier had better methodology, but that is a pet argument for me). To my mind, outrageous arguments such as the ones suggested above tarnish what could otherwise be very good work. The same is true of primary sources and ancient secondary sources, though in the latter case they often suffer from source problems as well and make do with comments such as “I don’t know how they did it, but here is what should have been done or what I would have done.”
As I do not need them for any sources (and even in Alexander studies, I find Tarn usually of secondary import), I can be somewhat amused at their eccentricities, but it goes to show that historians must always be vigilant lest their sources lead them astray. Of course this doesn’t mean that simple narrative with no explanation or analysis takes place, but the historian should make up their own mind rather than allowing the biases of their sources play a prominent role.