Just like my old physics teacher. Heard stories about him telling the students, that Pd (Palladium) is named after him (his last name had the same abbreviation).
Also jokingly using the screen of a calculator as a scale for weighing metal ball bearings.
But if you read a primary source, that's one persom who had the opportunity to make stuff up. With a secondary source, even if the primary it's based on is legit, there's some other guy who wasn't there and may either be lying to you or misinterpreting the primary source his report is based on. Each new level of isolation adds another opportunity to stack both lies and mistakes onto the data.
It's not that you can't go wrong with primary sources. It's that you can go a lot wronger without them.
Counterargument, secondary sources are often a good filter for bogus primary sources. This is the primary reason Wikipedia does not allow primary source references.
That's very different. Wikipedia doesn't allow people to edit their own pages. They don't have rules against linking to interviews with persons involved in an event, for example.
The main problem with primary sources is that they are often involved in the event itself - or at least greatly affected by it - which makes them the most biased.