Just from the details provided here, limited as they are, this study immediately raises some red flags with regard to its methodology. First of all, the participants were asked how much coffee they intended to drink in a day. They may not have actually drank that amount of coffee. People's intentions are often quite different from their actions, and some participants may have changed their coffee drinking habits over time, consuming either more or less of the beverage. The study does not seem to take this possible variation into consideration. It would have been much more useful to measure how much coffee people actually drank rather than how much they intended to drink.
Second, the organizers of the study deliberately limited their participant group to exclude people with significant health issues. Apparently, someone along the way decided that coffee probably couldn't do them much good anyway.
Third, there may well be a confusion between cause and correlation in this study. We may wonder whether coffee actually causes a lower risk of death or whether its consumption is merely present as a correlation. There may be confounding variables at work, variables like age, activity levels, overall health, and eating habits, that could make as much (or likely more) difference as coffee drinking in participants' lives. Yet the study does not account for these.
The conclusion about women coffee drinkers and ovarian cancer also appears shaky. Again, confounding variables like age, activity, eating habits, and health issues are certainly present yet not accounted for.
Overall, the study's results appear to be questionable. They claim too much on too little proof with not enough attention given to other aspects of the participants' lives and health.