Why is performing research that could eventually lead to a bedside test unrealistic? Genotypic markers are used in oncology to direct treatment all the time. Speed is not a particularly difficult hurdle to overcome, if clinical decision-making could be substantially improved.
1. We are developing ways to increase the availability and utility of genetic testing data at the bedside, and by "we" I mean me and people like me. This is something I actually know quite a bit about, and while we aren't there yet, we aren't going to get any closer by sitting around.
2. Searching for genetic variants associated with ischemic and hemorrhagic stroke endophenotypes is not only useful from a genetic standpoint, but also can help to identify new pathways that inform us about the underlying biology of these diseases. These searches can take the form of hypothesis-informed or completely agnostic strategies such as GWAS or exome/whole genome sequencing.
3. If you are up on the recent RFAs from NINDS, which it is my job to be, you may find that "acute stroke treatment" is not the only big push that Drs. Landis and Koroshetz are interested in fostering. Recovery and outcome is a big deal, with enormous associated healtcare costs, and genetic association testing has great potential in advancing the search for both novel therapeutic agents as well as valuable prognostic information for physicians and families.
I would humbly suggest that perhaps a more open-minded approach to cerebrovascular research might be valuable. Just because a project's goals aren't in line with your own does not mean that a) the project is not tractable or useful, or b) that it won't merit a higher priority score than yours. But congrats on going to ISC.