A flown topo won't help you with what's under the water surface in the channel, and if it's densely wooded and a narrow channel and flown in the summer you may miss some bank detail... however, in a flooded stream, the majority of the flow is on the overbanks so you will need a detailed survey of your overbanks, which on a large area is prohibitively expensive to be done by a field survey crew.
As cvg mentions, you do need the structures to be picked up, and depending on budget and your tolerance for error, a detailed survey of the bank/channel, though again, especially with possible debris in the channel, the fraction of water flowing in the channel during a flood is small and "holes" in the channel, though contributing some turbulence, are ineffective flow areas, and not necessarily worth the time and effort of survey.
I suggest that you use the 5' topo to select your cross-section locations so that your field survey crew has an idea where to go; cross-sections should not be straight lines, but rather perpendicular to the flow direction. What do you do when HEC-RAS gives you an error indicating that more x-sections are needed? Do you send the crew out again or interpolate x-sections? If the latter, wouldn't a 2' or 1' flown topo be more accurate?