I think that this would be highly unlikely. If I were were bothering to spend the money to engineer something, it certainly wouldn't have the characteristics of this virus. Not nearly deadly enough to make a difference if you are deploying it against an enemy. Kills the wrong people if you wanted to destabilize via panic - try killing young adults or children for maximal effect here. The biggest reason, though, is that most bioweapons have a really crappy track record. If virulent enough to achieve your goals, most viruses will just keep going right past your intended target to circle back and bite you in the butt. Most bacteria haven't panned out. Anthrax is interesting given the extremely difficult-to-achieve aerosolized spore delivery option, but relatively easily countered with antibiotics. Most US interest dried up years and years ago due to the inherent limitations of biological agents as weapons. Nerve agents were a much more sure thing.
Our knowledge of viruses only extends as far as our modern experience. The degree of human expansion into previously uninhabited areas, abnormal environmental conditions or the search for food or animals for research have periodically put humans into contact with reservoir species, leading to Ebola, Marburg, Machupo, Hantavirus, SARS, MERS. I'd put my money on this being a similar zoonotic exposure.
The long incubation period and extremely variable severity (though apparently with the ability to achieve asymptomatic transmission) does make this virus a lot harder to spot, as many won't even seek care. That, and the initial sweeping under the rug of severe cases by local authorities as far back as late November or early December, is likely what allowed this to spread so widely. I'm not yet convinced by the scattered reports of reinfection. Definitely possible given some chance of incomplete immunity, but as a mainly lower respiratory virus, it may simply be a case of nasal swabs being transiently negative when a BAL would have picked up virus particles.
At the moment, from my perspective as an ED doc, the most difficult thing about controlling the spread of this virus here is that it is clinically indistinguishable from flu or other respiratory viruses and we still don't have short-turnaround molecular testing available to identify it while the patient is still in the ED. As of friday night, I was still shut down by the health department on doing a swab of a university student with an unexplained fever and negative PCR respiratory viral panel (for the standard pediatric respiratory viruses, but doesn't include COVID) but suggestive symptoms, and we have already had (since the end of January) 3 positives as of last friday and one case of community transmission IN MY MEDIUM-SIZED COMMUNITY HOSPITAL. That's not even considering the cases with suggestive symptoms that CDC didn't authorize testing on due to geographic limits of testing or lack of sufficient severity of illness. Being told to self-quarantine because you maybe have something is one thing. Being told to self-quarantine because you definitely have the virus and might kill your grandparents if you don't is another entirely. I'm not sold on the efficacy of the no-test self-quarantine model.
As a result, I think that due to clinically silent or insignificant cases combined with rigorous requirements for testing (probably due to lack of availability of test kits for the last month) we have accidentally recapitulated the Chinese model of failing to achieve local control when we had the chance. I'm not looking forward to the next couple of months in the ED. We've already had about 20 staff quarantined, and that's only going to increase until we get real-time diagnostics in our hospital lab, which is still probably a few weeks in the future, if it happens at all.
By the way, I had a conversation about SARS with colleagues in late January and found this article. Really interesting. One lucky thing about this outbreak is that it doesn't seem that super spreading is as prevalent.
SARS article