That's literally the desired effect. What you're thiking of is war.
> It sounds like you could infect thousands/millions
Its not the right way to infect millions of people. It does not scale. Therefore not effective.
"The unlawful use of violence and intimidation, especially against civilians, in the pursuit of political aims" is a definition of terrorism. Death count is irrelevant (outside of spreading fear itself). Zero deaths in a terrorist attack could still be a very effective attack. Spreading something that causes a hideous disease in a city and causing mass hysteria could be an excellent terrorist tactic. Likely better than a bomb that kills a few people.
From this point of view, mailing of anthrax in 2001 was very inefficient.
Numerous USA calamities make a lot more sense when we study them with an eye toward who benefits rather than who is blamed.
The anthrax mailing caused changes and fear in US society with very little work. That sounds like a very efficient terrorist attack IMO.
For prions specifically, the main "problem" (which is a good thing overall for us) is that they're naked proteins, and they denature relatively quickly when outside of a protective environment. Spray them onto crops? Crops and soil are covered in small bugs and microbes which break down anything too long, including proteins and carbohydrate chains. No, in order to have prions in the food supply, you need something like cannabalistic cow-raising practices or other ways of recycling prions in a wet+hot safe environment.
How does this jive with the seemingly extreme difficulty of effective sterilization of prion-contaminated objects?
That is not how terror is defined though. The amount affected by terrorism is the amount of people that get terrorised by the act, not the amount that gets sick, dies or know someone who did. Using your definition 9/11 only affected aprox. 10.000 people but in reality if cased mass hysteria and fear in the US (and elsewhere) that is still there today.