I took it that the focus on the thoracic duct in this animal model was because this was the lymphatic drainage most likely to be 'sweeping' up from their cecal puncture and ligature model for sepsis.
I would presume if they used some other model (of the periphery, etc) --- or in humans with a diffuse inflammatory condition, that the drainage of choice might be different.
I'd agree completely that (if you give credence to the PAR1 paper) eliminating the inflammatory mediators at the lymph nodes would be preferrable to trying to play catch up once they are made.
Off to go ice climbing tomorrow. Keep an eye on the goats for me!