Graphene oxide will be better as the chemical groups attached to the GO may interact with gases in particular way. However, if it is based on pure resistivity graphene will be better.

Based on my opinion, following questions should be considered to reach the best results.

1. what kind of signal are you going to detect, electrical or spectroscopic? rGO yields better electrical signals than GO.

2. mechanism behind sensing, i.e. molecular recognition. GO might provide more opportunities than rGO due to abundant functional groups available. 

3. selectivity in sensing. sensors without selectivity are useless at all. I would recommend designing functional groups on either GO or rGO to match materials you are going to detect.

Hopefully, it helps.

Best regards,

Yunshen Zhou

What I understand is that your material will be composite of metal oxide and carbon for gas sensing.

Two key points to be considered.

1. making composite of metal oxide and carbon. 

As GO has huge functionalities, it will be suitable  than rGO to synthesize composite with metal oxide. More interfacial interactions with metal oxide and GO will be present in the composite. Whereas, in case of rGO, less interactions or interfaces will be present in the composite. Again GO will have better interactions with the gas molecules. 

2. If metal oxide is the active material for gas sensing, and  carbon material does not have direct role on gas sensing, then rGO will be of good choice as it has better electrical conductivity.