I'm having trouble picturing the problem. Are you trying to have a bullet leave multiple decals in it's path? Or decal non-colliders?
Maybe you can do something with collision layers and assigning the colliding objects to groups that determine the texture of the decal?
It sounds like you are already doing a great job and taking the right steps. The challenges you are facing aren't uncommon.
When it comes to non-functionals (response times, scaling, security, cost, update speed), you need to quantify them. Bring the stakeholders and your team together and set standards. If stakeholders have unrealistic requirements, this gives the team a chance to negotiate. If your team repeatedly fails to meet the standards they agreed upon, you have a more difficult conversation.
As an engineering manager, your job shouldn't be to know everything and predict the future. Your job is to use your technical intuition to create clarity and keep people accountable. Manage expectations and filter out the bunk between business and technology.
I hope you are doing well, it sounds like you are in a tough spot, but you can recognize it :) It's important to remember that it's possible your employer has unrealistic expectations of your role. Then you should talk to them about it. Expectations can be managed. If that is also not possible, it's okay to look for a better place. We don't have the time on this earth to take the heat for things we can't control.
I like refit too! It's super nice, but very quickly i realised the API does some weird undocumented stuff and isn't very consistently implemented. I feel more comfortable doing the boilerplate my myself (chatgpt helps) while i figure out the funk.
Hey, if the answer is: there's no obvious solution, than at least i don't have to feel too bad 😂
I'm sure the hardcore variant would have its uses. But the goal isn't necessarily to make fraud impossible, just evident. So probably more towards the latter option. And you are correct that you don't need a blockchain to create a distributed database that enforces consensus. It's just a neat tool you could use that scales pretty well, is relatively low maintenance(SWE hours not GPU hours), can adapt to a lot of cases, and is affordable for small and mid-sized companies. You could do the same by broadcasting your events to all your peers and having each peer save everyone's events to compare notes later. But this would be a hassle to setup and keep consistent.