By one-shot, do you mean an adventure or encounter that can be conducted entirely over a single online session, maybe three to six hours; or do you mean a single adventure, even if it takes more than one online session to accomplish?
This topic is of interest to me because that's what I'm engaged in right now, in both ways. I'm three sessions into a single adventure out of a projected total of twelve sessions; and I've been running, or trying to run, micro-adventures that can be concluded in a single three hour chat.
For the single, multi-session adventure, I'd say there is no particular obstacle beyond what everyone encounters in their medium of choice: player recruitment and attrition, scheduling, continuity, that sort of thing. Realistically, many "campaigns" don't last to the conclusion of their first adventure, anyway. Some players only want to play in campaigns and won't consider standalone adventures, but that's just one of any number of idiosyncratic reasons players will reject games.
The one-session adventure changes the nature of the challenge in several ways, though. Now you have to really pay attention to exactly how long you expect your session to run and how much you can truly expect to accomplish in that time. This will depend on several factors: rules system, preparation, the affordances of the medium, and the biggest factor that you have the least control over: your players.
In most systems, there are some parts of a role-playing session that will be more mechanically-intensive than other parts. Usually this means combat. When things go well in a three-hour session, I can fit in two mechanically-intensive segments interspersed between two less-intensive ones. A settling-in period and a wrap-up serve as bookends. Most GMs are more interested in the wrap-up and would prefer to dispense with the settling-in, but player dynamics rarely allows this to occur, so it's wise to plan for both. At the same time, if the GM does want some kind of denouement, it's important not to get carried away, or players will start popping off arbitrarily.
So that's when things go well. At their worst, which would be the one-session game I tried to run most recently, you don't even get through the first intensive encounter. If the players are interested in reconvening, that's great. (I'm currently in the midst of trying to determine whether that's true in my case.) Otherwise, you just have to be willing to scrap it and either try again later, or try something else later. At least, unlike a campaign, there's not as much investment lost if you decide to move on.
I could go on and talk about limitations of scope in these standalone sessions, but I think that's fairly clear based on what's laid out above. More than one red herring or plot twist, or often enough, even a single such curve ball, will usually guarantee that the session won't reach its resolution. Fast-paced, decisive action, with little in the way of mystery, is usually the way to do it. Perhaps you can have a single mystery take up most of the session until the players solve it, and follow that up with a brief climax and simple denouement, but you will certainly need players who can be satisfied with something like that.