Prior standard errors of shocks' standard errors

Dear Johannes,
First thank you for your previous guidance, I am grateful.
I am wondering that when should prior standard errors of shocks’ standard errors be set to infinity, when should it be set to 2?
For instance, in this paper DSGEmodelofStockBubblesandBusinessCycles.pdf (522.6 KB) published in Quantitative Economics, on page 21 (top right original page 619), prior standard errors of shocks’ standard errors are set to infinity instead of 2 as in some other literature.
Thank you very much and look forward to hearing from you.
Best regards,
Jesse

A prior is always subjective. The best way to see whether there is a big difference in terms of implications is to plot the resulting distribution.