All of the above badly fail Occam's razor. That's why it's a paradox:
You can't fail the Occam's razor, it's not a rule that's always true, just a rule to use when you have no other. Every scientific discovery we have ever made has failed the Occam's razor at the time. Badly. If they didn't, there would be no use for science.
No matter how unlikely those scenarios are, each of them is decreasing the probability of the limited choice that either aliens aren't there or we would have met them. It is a false dichotomy – all the insignificant scenarios add up to something significant enough to not be ignored as an option.
There are millions of reasons why we wouldn't meet aliens even if they were all over the place, from insufficient telescope resolution to looking at the wrong places to pure chance.
One thing that would seem reasonable to expect is that if two civilizations became aware of each other, one of them would go extinct or one of them will push the other ahead, either by giving them technology directly or at least by inspiring them, thus transforming them into a more advanced one. The latter option might mean that the two civilizations merge, at least technologically.
That's kinda cool and funny, because under those assumptions there is a uncanny similarity between the expectation to be born in a civilization that has met aliens and a civilization that is in its middle age. Not only both assumptions are a result of your expectation to be average (and be outran by those aliens that you'd be meeting), but actually meeting them could transform you from a toddler stage to a middle aged one in a few centuries.
Both even seem logically equivalent, which means that we should expect to be extinct in a few hundred thousand years for much the same reason we expect that aliens are rare.
Valid conclusions actually, as long as the probability of experiencing each moment in the history of the universe is the same or very similar. What if the universe becomes more unlikely with each day, though? Say, for example that the multi-world interpretation of quantum mechanics is “true” and most “timelines” suffer from some kind of catastrophic event that leave the interesting ones more and more unlikely? Anything like this would be completely unobservable and unmeasurable by us, and I'm not completely convinced that it would be the more complex option that necessarily violates Occam's razor – we know nothing about the universe as an object, and we most likely never will.
100 million years is more than enough time to colonize the galaxy with ships only able of 0,1 lightspeed.
What are the energy requirements for that? Is the time necessary to rebuild the civilization at the next star system taken into account, which at 0.1 lightspeed is probably a couple of orders of magnitude larger than the travel time, and speeding it up would further increase the energy requirements for the trip?