As long as the current exchange rate mechanism (which is currently responsible for the vast majority of NNS proposals) is in place, the reject fee should not be increased any further.
The reason is that it is always possible for exchange rate proposals to be rejected (usually, due to network issues, preventing the conversion rate providers from casting their votes in time). When the conversion rate provider neurons run out of ICP, the ICP/XDR rate will no longer be up-to-date until the they are topped up.
I agree with those who are in favor of implementing a more permanent solution to the spam problem instead. Personally, I think the first step should be the “voter-based” rewards mechanism discussed here.
If the reject fee is to be increased anyway to prevent advertisement spam, this should be done after the new exchange rate mechanism is deployed.