Subcutaneous dosing of large molecules is basically a method of creating a slow(er) released depot of material. For molecules with a long half-life, like Mabs, this is ideal in terms of pharmacokinetics. The alternative, slow infusion intravenously, is uncomfortable for the patient as they must travel to the infusion center/doctors office and spend up to several hours there.
I don't believe there is a standard sub cutaneous formulation. Single bolus sub cu injection volume is limited to ~1 ml. Therefore, the solubility and viscosity of the product are critical. Aggregates can cause immune responses and must be avoided. Extremes of pH will cause irritation. Pre-filled syringes for self-injection are pretty standard now. So, selecting leads that allow a viable sub cu formulation early on is probably a good strategy.
New wearable auto-injection devices may be able to elimnate some of the volume limitation while providing convenience.
If on the other hand, the treatment is required immediately to treat an acute situation, such septic shock, slow release is obviously not the answer.

I would say that the injection system is also dependant on the drug itself. Even if the product shall be used in IV, it is soemtimes better to inject in SC or IM because the injection shall be done either by using a DM for self injection either by a nurse or by a trained person. But it is impossible to consider the injection of a large volume in IM or SC as this shall be very painful. So if your drug cannot be concentrated in the formulation and you need to inject more than 10 ml, IV shall be of interest for you better than other type of injection.
So at the end, it is not something that you can easily design prior to initiate all the development of the formulation. In addition, this shall also be driven by the shelf life of the product which might need to be freeze dried, so then the formulation and the reconstitution volume shall be considered.

It strikes me that the questions -
"Is there a consensus on formulation?
Competitive advantages?
Is there a real benefit for iv, beyond compliance?" are inappropriate in the absence of specifics about the drug of choice, indication, etc.
For example, benefit of iv can be determined for a given drug, but not in general, without focusing on a specific drug. Drug body distribution depends on the drug properties, and also on the route of administration; not considering this together makes little sense.

We seem to agree. By giving an example of a specific drug - remicade - you illustrate my point. General rules may be developed based on experience, for example:"For drugs having properties such as remicade, general rule is ... etc.