The paper investigates the performance of integrated sensing and communication (ISAC) networks, focusing on balancing sensing and communication (S&C) performance at the network level. It proposes a cooperative ISAC scheme that utilizes multi-point (CoMP) coordinated joint transmission and distributed multiple-input multiple-output (MIMO) radar techniques to enhance S&C services. Stochastic geometry is employed to characterize S&C performance, revealing key cooperative dependencies and optimizing network parameters. The Cramer-Rao lower bound (CRLB) expression for localization accuracy shows that deploying $N$ ISAC transceivers enhances average cooperative sensing performance according to the $h^2 N$ scaling law. However, this scaling law is less pronounced compared to the $N^2$ performance enhancement when transceivers are equidistant from the target due to path loss from distant base stations. The paper also derives a tight expression for communication rate and presents an algorithm to determine the optimal cooperative cluster size. Based on these expressions, an optimization problem is formulated to maximize network performance in terms of two joint S&C metrics. Simulation results demonstrate that the proposed cooperative ISAC scheme improves average data rate and reduces CRLB compared to conventional time-sharing and non-cooperative schemes, achieving a better S&C performance tradeoff at the network level.The paper investigates the performance of integrated sensing and communication (ISAC) networks, focusing on balancing sensing and communication (S&C) performance at the network level. It proposes a cooperative ISAC scheme that utilizes multi-point (CoMP) coordinated joint transmission and distributed multiple-input multiple-output (MIMO) radar techniques to enhance S&C services. Stochastic geometry is employed to characterize S&C performance, revealing key cooperative dependencies and optimizing network parameters. The Cramer-Rao lower bound (CRLB) expression for localization accuracy shows that deploying $N$ ISAC transceivers enhances average cooperative sensing performance according to the $h^2 N$ scaling law. However, this scaling law is less pronounced compared to the $N^2$ performance enhancement when transceivers are equidistant from the target due to path loss from distant base stations. The paper also derives a tight expression for communication rate and presents an algorithm to determine the optimal cooperative cluster size. Based on these expressions, an optimization problem is formulated to maximize network performance in terms of two joint S&C metrics. Simulation results demonstrate that the proposed cooperative ISAC scheme improves average data rate and reduces CRLB compared to conventional time-sharing and non-cooperative schemes, achieving a better S&C performance tradeoff at the network level.