China is absolutely NOT an interesting place from the perspective of government contracts, particularly for US citizens that don't want to end up in jail, however. Anything defense related will fall under ITAR and associated tech export bans (this includes just working on these programs, not just physical movement of goods).
Additionally, if you're paid in Chinese currency it might as well be monopoly money to a foreigner.
There's also the massive ethical dilemma of supporting a government that is actively engaged in ethnic cleansing (regardless of the whataboutism).
1. Companies that made a business out of doing SBIR grants (I think the previous poster may have meant the same with SBIR farms). They write one grant after another and do some BS research, write reports and in the end, trash the results and move to the next SBIR
2. Small Start-up that a basically linked to big military suppliers and since the big supplier is ineligible to appply, they apply for them and among the big big stack of applicants, they are selected for the SBIR. Must be really superior ideas I guess. Or huge luck :-)
The SBIR grant was designed for neither, but blue sky ideas that VC don't want to fund. My experience with SBIR has been terrible. Especially
A. Idiot reviewers. I can only tell you that an idiot and google are a very dangerous combination.
B. Idiot management. One lady that is in charge of giving out the SBIR has emailed me that she has occasionally given SBIR to inexperienced (speak not having done an SBIR before) researchers. So she basically handles the grants to companies mentioned above under "1", the SBIR farms. She has no idea what she is doing. SBIR was not designed for that.
In my opinion, and I may be wrong, SBIR were designed to be open to the best blue sky ideas that may be too risky or too early stage for regular VC. This already excludes the existence of SBIR farms.
This was kind of my experience, though it isn't quite so malicious generally. The overhead in setting up the contract, while not ridiculous, is enough that once you get one, you usually end up getting another, and so on until you become an SBIR farm. The grants come in phases, Phase 1 being research and proposal, Phase 2 prototype, and Phase 3 basically being follow on contracts for production/continued development. You have to get them in order and if you only farm Phase 1 contracts you won't get super far. That said getting past Phase 2 is quite hard as the kind of stuff they generally ask for (weird one off fixes for various government procurement requirements) don't generally commercialize well.
They're frequently moonshot type efforts for the size of businesses they're targeting and the ROI is low unless you're using them to build and pay a team for some other goal. That said, if your goal is to build a team/small facility and you know what you're doing, it's not a bad option. It just isn't really anything like what the rest of the software world's VC funding model.