"I don't think you understand what "co-evolution" means. It has nothing whatever to do with the inheritance of acquired characteristics, or even (directly) with inheritance at all."
You're missing the point. If co-evolution is happening within a hundred years, it must be Lamarckian, because otherwise the "random mutation" part would not have time to catch up.
"It simply means that two species have strong and mutual environmental impacts on each other, such that if one species evolves some new or improved capability, the other will be under an immediate selective pressure to respond in some way, and so their evolution is linked by a feedback loop via natural selection."
Exactly, but if this process can happen within a measley few generations, then the adaption process is not random in order for evolution to occur at that pace.
Stultis, you want to take this, or do you want me to?
You don't usually need to wait on random mutation. Most populations maintain a significant amount of variation at any one time, and thus can respond, often on time scales of a few years or less, to sufficiently extreme selective pressures. See, for instance, the book The Beak of the Finch which discusses documented, and nearly instantaneous, selective responses to droughts in the Galapagos Islands. (I.e. the results of the droughts were such that the sizes and natures of seeds available to the birds shifted significantly, and so did the average beak sizes. There was already a range of available beak sizes in the population, so differential survival rapidly shifted the average or typical beak size.)