The performance of electronic systems for radio-frequency (RF) spectrum analysis is critical for agile radar and communications systems, ISR (intelligence, surveillance, and reconnaissance) operations in challenging electromagnetic (EM) environments, and EM-environment situational awareness. While considerable progress has been made in size, weight, and power (SWaP) and performance metrics in conventional RF technology platforms, fundamental limits make continued improvements increasingly difficult. Alternatively, we propose employing cascaded transduction processes in a chip-scale nano-optomechanical system (NOMS) to achieve a spectral sensor with exceptional signal-linearity, high dynamic range, narrow spectral resolution and ultra-fast sweep times. By leveraging the optimal capabilities of photons and phonons, the system we pursue in this work has performance metrics scalable well beyond the fundamental limitations inherent to all electronic systems. In our device architecture, information processing is performed on wide-bandwidth RF-modulated optical signals by photon-mediated phononic transduction of the modulation to the acoustical-domain for narrow-band filtering, and then back to the optical-domain by phonon-mediated phase modulation (the reverse process). Here, we rely on photonics to efficiently distribute signals for parallel processing, and on phononics for effective and flexible RF-frequency manipulation. This technology is used to create RF-filters that are insensitive to the optical wavelength, with wide center frequency bandwidth selectivity (1-100GHz), ultra-narrow filter bandwidth (1-100MHz), and high dynamic range (70dB), which we will present. Additionally, using this filter as a building block, we will discuss current results and progress toward demonstrating a multichannel-filter with a bandwidth of < 10MHz per channel, while minimizing cumulative optical/acoustic/optical transduced insertion-loss to ideally < 10dB. These proposed metric represent significant improvements over RF-platforms.