No, HIV could not become pulmonary, spread by coughing and sneezing. Its RNA is vastly different from the viruses that can. Influenza, however, as a class of viruses, is special.
To start with, it has a large number of what are called “flexible genes”, genes that are prone to mutation. This is why there are constantly new iterations of the flu, even in the same flu season. To show how radical this is for H5N1, here is a pdf chart of its known variations over time:
http://www.who.int/influenza/gisrs_laboratory/201101_h5n1evoconceptualdiagram.pdf
To explain, the “H” in H5N1 is the means by which the virus enters cells, and the “N” is the means by which after reproducing millions of copies, they all bust out of the dead cell. There are 17 different kinds of “H” factors, which had been H1, H2, and H3, which people have been exposed to for many years, and so have at least partial immunity. But H5 is novel, so no humans are immune to it, even partially.
There are some 137 types of “N” factor, though just a few apply to humans. All the antiviral drugs such as Tamiflu focus on inhibiting the “N” factor.
Before I mentioned that H5N1 only needs five mutations to be easily transmitted between humans. Because it has oddly and stubbornly maintained its extraordinary 60% mortality rate among those infected, this is extremely bad.
The chief epidemiologist of Vietnam, internationally respected, made the astounding discovery of a herd of pigs, each of which had on average five distinct mutations of H5N1 in their bodies. These mutations were in competition with each other, with the survivor of the “semifinals” competing with other semifinalists from other pigs, until a “champion” mutation infected the whole herd.
From this he asserted that the virus is using a vast number of herds and flocks of animals like a slow but enormous computer, to develop superior mutations.
The ferret has perhaps the closest immune system to humans in the animal kingdom. For this reason, they are regarded as a “canary in a coal mine” for deadly human epidemics.
A research scientists took ferrets, infected them with H5N1, and used them to selectively cross breed the virus with that of H1N1, with known virulence to humans, just by being in their bodies at the same time. It killed off his ferrets in record time.
This is how we learned of the 5 needed mutations, which in effect, happened “naturally”. However, the scientific community and government intervened to demand that when his work was published, that there should be censorship of the exact process, as it was too dangerous to publish.
The inability of wild-type HIV to spread by respiratory droplet has nothing to do with being an RNA virus. It has everything to do with whether or not the virus can avoid desiccation and stay infectious/viable when aerosolized, and whether there are receptors for HIV on the cells in the airways. Growing lentiviruses in cells that express envelope proteins with tropism for lung epithelial cells would significantly increase their ability to bind and enter these cells. If one wanted to weaponize HIV, changing its tropism is one way you could do it, God forbid.