The best examples of far reaching global pandemics are the Black Plague and the 1918 influenza epidemic. Apart from not having the kind of lethality rate you're looking for, both examples were able to spread widely due both to a lack of any effective means of controlling travel, and a widespread lack of effective medical care.
If you're envisioning a modern pandemic then, your disease needs to spread faster than it can be quarantined, and be widespread before alarmed governments can shut down international travel and quarantine impacted communities.
As far as your first two criteria are concerned, you can just look at influenza. It's airborne and incredibly contagious. For the third and fourth criteria, you have a couple options.
Option 1 is that the disease has a moderate incubation period, say 4-8 days, and a gradual decline during which the person thinks they just have a normal flu that hangs around for a week or two, during which they're still actively moving around and infecting people, before they start getting REALLY sick and eventually die.
Option 2 is a unusually long incubation period of more like 10-20 days, after which the infected individual becomes increasingly symptomatic over days and dies very quickly.
Option 1 has the greatest potential for distribution in that by the time anybody realizes there's an epidemic occuring, the infected percentage of the population is VERY high. However, it also allows the most time for response and treatment once the global medical community becomes aware of what's going on.
Option 2 allows less overall time for distribution, but when people started getting sick, there would be millions and millions of dead all at once, and with a spread pattern like this, medical professionals would be most at risk of infection and death before any countermeasures could be devised, which also addresses your fifth item. If it's not susceptible to standard antibiotics/antivirals/antifungals, then under these conditions by the time someone figured out (or developed) something that DOES work, it would be too late.
I'll also point out that the actual disease doesn't have to be what kills EVERYBODY. Throughout the developed world at least, the way most people live relies on a complex and rather fragile economic system that ensures that food and clean water and electricity and so forth get to everybody. If you had enough people die quickly enough from the virus (say 30-40% of the population), that would conceivably disrupt things so badly as to cause catastrophic collapse of industrialized society almost overnight, which means that the people who didn't die immediately from your superflu or whatever would still die anyway from violence, starvation, and the more normal diseases associated with having a few million dead bodies lying around with nobody to handle it.