We haven't yet collected evidence that says terrestrial life originating on Earth is impossible or even unlikely, nor has anyone produced an alternative explanation that doesn't involve introducing additional external factors. It may indeed turn out to be incorrect (although I wouldn't be on it), but it's in no way comparable to geocentrism which simply couldn't be reconciled with experimental observation.
If you lived on Long Island, and didn't know where humans came from, but you knew about how big the earth was, would it be rational or in accordance with Occam's razor to assume homo sapiens originated in your vicinity? Most people would in fact be wrong if they even assumed humans originated on their continent.
I feel like people get sidetracked into talking about evidence, when the issue is that life originating on Earth should not be the default assumption. I feel like someone more lucid than me could make a good Bayesian sort of argument.
It's not exactly the default assumption though. Historically the default is that life was magicked into existence or came here from a previous world or some other creation story. The current abiogenisis story of molecules->proteins->cells->Life is relatively recent.
The other thing is that because we have a fairly... ?linear record of evolution: if life did originate somewhere else it got to Earth in something like prokaryotic single cell form 3.5 billion years ago. Which means we can't use the idea of panspermia to tell us anything. It's neither predictive or explanatory. It tells us nothing of how or where life did originate or how life got started on Earth or what or where we should look for or expect in the rest of the Universe.
The current theory suggests life could pop up extremely quickly, which itself already implies life is likely to exist in the rest of universe on some level so panspermia doesn't even mean anything with regards to how likely life is anywhere else.
The difference is that you've never seen evidence of humans living anywhere outside of Long Island.
Clearly if you see humans living all over the globe they almost certainly originated elsewhere. But if humans only live on LI, then it's reasonable to conclude they originated on LI.
It's an intuition or speculation at this point. The only major evidence we have is that a logarithmic graph of information vs time in evolution seems to cut off abruptly at the age of the Earth, suggesting that the process started before Earth if a constant rate is assumed.
This is definitely not conclusive as it could also mean that some other process with a different rate or probability structure started life. You can't always just do a regression and draw a line and assume that it holds for all time, as many startups discover. ;)
We also know that organic molecules are ubiquitous in the universe, so there's that too.
Given the gaps in our knowledge of early life, a definite statement in either direction is unlikely to be in the offing. However, terrestrial life is decidedly unlikely. Life is decidedly unlikely, and no one has so far suggested an even remotely plausible mechanism for abiogenesis on Earth. This means if you are betting, an extraterrestrial origin is the safe bet; if the odds of abiogenesis are so slim, while life once formed is tenacious and hardy enough to travel through space, it is more likely life traveled here.
It's not clear to me how much panspermia really helps with the abiogenesis problem.
Panspermia might increase the window of time and set of planets where the abiogenesis of Earth's life could have occurred. But are there any grounded estimates of the factor by which it expands those windows? For example, is it safe to assume that if abiogenesis occurred on any planet in the Milky Way before the Earth formed, then samples of that life would have reached Earth within say 200M years after the Earth formed? That would require estimating the rates at which samples leave the planetary gravity well, how likely those samples are to reach Earth's gravity well, how likely the samples are to survive the trip, etc. I have not yet read any detailed estimates of this kind (but I'd like to!).
Another problem ... if the answer to the previous question is "yes", then that might expand the abiogenesis window by, say, 10^13 (say 200B planets and a 50-fold increase in the available time window, ignoring other galaxies). Is that enough to flip abiogenesis from "not likely" to "likely"? 10^13 is a big number but if abiogenesis is extremely unlikely then we're already dealing with very low probabilities. In other words, panspermia only helps if the probability of abiogenesis occurring on a planet in a given time window falls in a certain range, and in context that range might be quite narrow.
We have a seemingly pretty firm age for the universe of 13.7 billion years; earth is 4.5 billion years old, and life on earth is at least 3.5 billion years old.
So I think the step of going from nothing to the first single celled life is implausibly large to happen in a few million to a billion years, and to me, Occam's razor would imply that something necessary was developing during the previous 9 billion years, such as the development of chemotrophs underground in deep space.
