I'm not informed enough on the topic to speak anything near authoritatively but there was an essay posted here ~6(?) months ago that spoke about how >2-3% inflation effectively serves as a means of effectively reducing government debt.
I would imagine your 5% number has to assume some baseline of inflation and that 5% would increase as inflation increases beyond said baseline.
It's true, you can gradually default on the government's obligations by changing their denomination to make the underlying promises worthless. That's consistent with my third hypothesized path.
I don't see any way that inflation jumps gently in that case. I'm suspecting we wouldn't jump straight to hyperinflation, but annual inflation 30-50% would be my guess for the first couple years that 0% of government services were covered by revenue.
I would imagine your 5% number has to assume some baseline of inflation and that 5% would increase as inflation increases beyond said baseline.