The answer to this question is certainly yes: diaries and journals can be very useful as primary sources. But I'd hesitate to answer the rest of the question. Historians are all looking for different things, and it's very likely that a journal that I find valuable would be useless to most other historians. Likewise, a journal of obvious historical value -- one by Abraham Lincoln, for example -- would be of no use to my work whatsoever. Generally we're looking for sources of this nature that are detailed, personal, and accurate -- but most of all they need to be relevant to be useful.

Yes! They can be helpful in providing insight from a personal level (the author of the diary, their family and friends, and their everyday interactions) all the way to a national or international level if they're writing about current events. While all sources are biased, diaries and journals are biased in different ways than other sources like newspapers or chronicles. Someone writing about large-scale events (wars, conflicts, assassinations, fights, etc.) might have less reason to distort the truth than a newspaper or a government. A hypothetical example: a student in Vienna protesting during the 1848 Revolution might write in their diary that they saw a friend get arrested for no reason, when the official report would probably say that the arrested individual was committing a crime. It's up to the historian to decide which is more likely/accurate (which is difficult if not impossible) but if there are enough personal accounts to counteract the official account, sometimes new truths can be discovered. This is all hypothetical and very fuzzy, but that's history for you.

As far as your second question goes, I'd agree that it's difficult to say. I think safe (and extremely general) qualities of a useful journal are that the journal can be accurately dated (so we can tell when the author lived and that they aren't pretending that they witnessed events which happened centuries earlier) and that the journal has some substance (more than a few entries). The larger the journal is, the easier it could be to understand the author and their biases.