Its like when some scientist reports they killed 70% of cancer cells in the lab and a news report comes out with scientists develop treatment for 70% of cancer.
You cannot just take "scientist reports they killed 70% of cancer cells in the lab" and then follow-up with a question of "what percent of cancer did the scientist cure?" and then use the only percentage you saw ("70%") as the answer. The information given wasn't not sufficient to answer the question asked; in a test scenario, sure make an assumption but in the real world just get clarification ...
I mean it's not though. The question is about "is soft" not "seems soft".
These slight of hands are pretty crucial to actually reading in the real world (or conversely, deceiving people).
It's like if you read a press release about how X is 120% better than Y. Judging from the press release, X is clearly better than Y but then if you actually do your own research you find for your workload X is actually 10% worse than Y.
What moisture level do crackers become soft?
I'm sorry, 5 percent isn't the correct answer; they actually get soft earlier than then.
This seems like one of those questions like "what's the difference between an abstract class and an interface in Java" where you're looking what the expected answer is as opposed to the actual answer. (w.r.t. Java; the answer they're looking for is that you can define a default method in an abstract class. It doesn't matter how long ago you could do that in an interface ...)