The span tables are based on the full section strength and does not consider local effects, and therefore you need to check the pipe under local support conditions additionally. My previous response was showing you that the local effect needs to be considered. I see that I needed to openly write this there. Uniform line support or local line support are going to be investigated as local effects, I agree that local line support will give you worse local effect, and reducing the span length will reduce local effects.
How are you going to warrant that all the support will act continuously under the full water conditions, especially on the pipe rack? Do you think secondary beams will have the same amount of support loads on the primary frame members? How are you going to warrant the same length line contact for all the supports?
The diameter to wall thickness ratio in your case is around 100. The pipe will have local stresses and radial displacement around the support locations. This needs to be investigated. The addition of internal pressure stresses on the pipe to the local stresses may cause yielding or sometimes fracture.
So you need to investigate the local effect with hand calculation or other methods to accept this pipe is adequately supported.
The best is to use large span length with adequate saddle support or a support which restraints the pipe's full circumferential at support locations for this type of piping/pipeline/duct. You did not mention of the code that you are suppose to use. However if you look at the water supply pipelines they all have large span length and supports that have reinforcement ring(s)to reduce the local effect. This way the pipeline is more economical.
I guess you will reinforce your assumption of "My understanding is main concern would be the bending stress (deflection). Since the deflection is minimal in this case (due to very little span), it`s not likely to fail." by a proper calculation and I hope the result will show that you are right.
Good luck.