That's a rather pedantic interpretation. To give one counterexample. Claims can have merit in terms of how the long term events that follow. I have certain ideas of what will follow from e.g. a given proposal; but that's often given many assumptions that went unstated in the original source. By knowing the direction of bias I can postulate what might come from the author's assumptions for those unknowns as well, and perhaps see an angle I hadn't or adjust my "internal bayesian function" of outcomes; which absolutely leads to a subjective judgement of merit. (but despite subjectivity is still to my eyes a best-we-can-do-in-absence-of-more-data question of merit; necessary due to the messiness of "real life")
To bring this very hand wavy generalization back to the topic, look at many arguments posed in the anti-marijuana camp (as discussed in this very writeup) being based on extremely pessimistic to the point of unrealistic assumptions in order to derive the conclusions they prescribed; conclusions in isolation that could be seen as terrifying but if you consider author bias can be accounted for more reasonably.
Do read the tail of my prior post. If only we were in a world where logical purism was possible you would be preaching to the choir, I promise you. But given a world with hilariously incomplete information, ESPECIALLY often when having to judge the merits of a decision, one must make a judgement that takes into account aspects outside of literal content. (If you're a CSer at all I would expect this is a very familiar concept from having to do anything from project estimation to architecture in the presence of unknowns)
(Sister post Linkregister says it better than I did, frankly)
Perhaps in a world where everyone always deals in facts and everyone only tells the truth that statement is true, but that is not our world.
People lie and distort facts for their own personal gain all the time.
See for example, the scientists paid by the cigarette companies that produced data showing that smoking was safe.
Knowing who said something (and knowing who is paying someone to say something) allows you to better calibrate your skepticism to see if what they are saying might be biased by their financial or other interests.
You should be equally skeptical of any scientists stating things like that. A scientist has all kinds of motivations to publish novel research. The first scientist publishing that smoking causes cancer had as many motivations to lie as the scientist that published research claiming it didn't. The latter was direct financial support, the former was prestige and fame that would go directly into a tenure package and be considered during grant approval.
If ad hominems become acceptable then you quickly open the door to worthless discussions around things like climate change because everyone can point out that the scientists behind the research are motivated to advance anthropogenic climate change or risk being ostracized.
Evaluate the content, not the messenger. Anything else is lazy.
If you are concerned the content is a lie, ask for the backing data, don't dismiss it.
To bring this very hand wavy generalization back to the topic, look at many arguments posed in the anti-marijuana camp (as discussed in this very writeup) being based on extremely pessimistic to the point of unrealistic assumptions in order to derive the conclusions they prescribed; conclusions in isolation that could be seen as terrifying but if you consider author bias can be accounted for more reasonably.