Coronagraphs are a promising solution for the next generation of exoplanet imaging instrumentation. While a coronagraph can have very good contrast and inner working angle performance, it is highly sensitive to optical aberrations. This necessitates a wavefront control system to correct aberrations within the telescope. The wavefront requirements and desired search area in a deformable mirror (DM) demand control of the electric field out to relatively high spatial frequencies. Conventional wisdom leads us to high stroke, high actuator density DMs that are capable of reaching these spatial frequencies on a single surface. Here we model a different architecture, where nearly every optical surface, powered or unpowered, is a controllable element. Rather than relying on one or two controllable surfaces for the success of the entire instrument the modeled instrument consists of a series of lower actuator count deformable mirrors to achieve the same result by leveraging the conjugate planes that exist in a coronagraphic instrument. To make such an instrument concept effective the imaging optics themselves must become precision deformable elements, akin to the deformable secondary mirrors at major telescope facilities. Such a DM does not exist commercially; all current DMs, while not necessarily incapable of carrying optical power, are manufactured with flat nominal surfaces. This simplifies control and manufacturing, but complicates their integration into an optical system because there is oftentimes a need to pack several into collimated space. Furthermore, high actuator count DMs cannot approximate low order shapes such as focus or tip-tilt without significant mid-spatial frequency residuals, which is not acceptable for a coronagraphic high-contrast imager. The ability to integrate the wavefront control system into the nominal coronagraphic optical train simplifies packaging, reduces cost and complexity, and increases optical throughput of any coronagraphic instrument. This adds redundancy, increases controllability of the complex aberrations, and mitigates both cost and risk associated with a single high-actuator count device that the entire instrument performance relies on. Here we simulate an optical system with a combination of controllable imaging optics both with and without a high order DM at the pupil. This example instrument is based loosely on the current DM technology being considered for the WFIRST CGI, and is merely an example of a larger trade study to be done to optimally balance actuator requirements, controllability, and wavefront quality. The relative performance of each configuration with regard to contrast, achievable bandwidth, and redundancy is discussed. The overall performance enhancements and risk associated with actuator failures on the assumed DM technology is also evaluated.