Linguistic and other evidence strongly suggests that the Polynesian islands were settled from west to east (probably originating from Taiwan) and Kontiki drifted with the currents from east to west.
What's interesting is that the Polynesians settled the outlying islands by sailing into the prevailing winds (making it easier to return home if the failed to find land).
They used efficient outrigger canoes which could be tacked or, sometimes even better, shunted, into the wind.
They also did a lot more than follow birds, they relied on keen observation of the ocean (interference patterns in the swells caused by an unseen island), the colour of the clouds (clouds are darker underneath when there is land beneath them) and relied on a sophisticated 'star compass' for latitude navigation.
If their home island was at latitude -18deg for example, they knew which zenith stars passed above their island, and which stars set at what point on the horizon at that latitude.
So after an eastward exploration run, when they wanted to find their way home, they just kept their islands's zenith star at the zenith and sailed west with the wind till they made landfall.
Polynesian non-instrument navigation is undergoing a revival. See:
Yes, the National Geographic article / series I mentioned nearby in this thread:
https://news.ycombinator.com/item?id=16922701
says those sorts of things. So does the Morris West book, IIRC (some of them). In fact, some of the photos in the NG article show a device that may be the same as the "star compass" you mention.
A significant part of our story has simply been washed away by rising sea levels, once the Ice Age ended.
Too funny: Kon-Tiki went from South America to the Polynesian and it's only more recent research which indicates it might have been the other way around.
What we realize now is how little we know.
During the minimal sea level, the peninsula of Asia included Sumatra, Borneo, and Java, while Australia included New Guinea, and the sea almost to (but not connected to) Timor. The minor islands of Indonesia (and Sulawesi) remained disconnected from either continent. The resulting gaps, however, would have been smaller: you could have crossed the straits without losing sight of land.
The Torres Strait is full of small islands bridging the 150km gap between PNG and Australia. Not only that, but much like the Strait of Dover or the Bering Sea it was most certainly a land bridge until the holocene, and it remains extremely shallow.
(If Wikipedia is right about the time of Tasmania's separation from the Australian mainland and Tasmania was already populated at the time, the feature would've had to have been preserved on both sides of the Bass Strait for about twice the time between modern English and Proto-Indo-European.)
It isn't possible to verify a claim like this (and therefore, it also isn't possible to honestly make a claim like this...) - there are plenty of traditions that claim to draw on 70,000 years of history, but no reason to believe that they actually do.
Some versions of the Sumerian king list document that a single king reigned for 36,000 years.
Just did a quick search:
https://www.google.co.in/search?q=national+geographic+articl...
https://blog.nationalgeographic.org/2014/03/03/hokulea-the-a...
https://news.nationalgeographic.com/news/2007/09/070927-poly...
Frankly, all of the examples have a click-bait tone and deal in those topics related to "secrets of the pyramids" cable shows: controversial until proven by an amateur; science in the dark until it asks the primitive people about one "simple algorithm;" surgeons use stone age tools; Greek objects on a Pacific island (but not really).
My debunking, in order:
Yes, we now know Polynesians were long-distance intentional sea-farers, but Kon-Tiki did not prove that and was very controversial (and unscientific) in itself. Polynesian-South American contact is hypothesized but doesn't have solid evidence (sounds more neutral than "isn't fully accepted"). Currently only the sweet potato and palm tree genomics show it was likely, still waiting for artifacts or human remains that could provide solid evidence.
About Polynesian sailing, others have covered the complexity and accomplishment of their navigation, I will add that they also likely travelled seasonally: go in one season, come back when the winds shift 6 months later. Recently, the crew of the Polynesian Voyaging Society sailed around the world using these ancient techniques.
There was likely a land-bridge across the Torres Straight at glacial maximum, but not across the Wallace line further east (see the Sundaland comments). Anyway, this comment is so vague as to not really say anything.
I get the idea that we modern industrial humans under-estimate what is possible with stone-age technology and ingenuity, but the link to modern scalpels is but one minor point (that they had sharp tools). You need a lot of other properties to build things, some cultures had them available and developed them (rope for binding, cloth making, etc). One thing to remember: when Captain Cook landed in Hawaii in 1778, the Hawaiians were essentially in the stone age (though some contact with Spanish galleons or the floating remains thereof is speculated). Very soon thereafter, an ambitious local chief used western advisors and purchased guns to conquer all other islands.
You say "Greco-Roman lamps turning up [in] Palau" but the article you linked shows that it is only the design of ceramic oil lamps in Palau that might be traced back to a 6th c. bronze Byzantine lamp found in Thailand. The article itself does much to play up the sensationalist nature of Greek influence in Micronesia, but draws lots of tenuous connections between a few pottery shapes. And both the article and your short summary both fail to mention the possibility of parallel development of a basic implement (only so many ways to burn oil with a wick in a vessel). All in all, the article reads more like Popular Science (with unrelated-topless-native-woman image) than a scientific paper.
The Polynesians were capable of voyaging across the Pacific to South America, and another Austronesian group made it all the way to Madagascar. This was pretty controversial until Kon-Tiki; even now, Polynesian-South American contact isn't fully accepted.
(IIRC, we weren't sure how the Polynesians even found so many islands, until an archaeologist asked a few of them and discovered that there's a simple algorithm for it: follow the birds.)
Australia had to have been sailed to from somewhere, possibly more than once, possibly intentionally.
Stone tools can be pretty good, especially obsidian. There are surgeons today who use obsidian knives.
Roger Blench has a paper on Greco-Roman lamps turning up as far afield as Palau, which implies more complex trade networks at an earlier stage than most people would think: http://www.academia.edu/9237516/The_Romans_in_Palau_how_did_...
And now, Neanderthals with boats. This isn't really new -- Homo floresiensis (the ebu gogo, maybe) had to have gotten there somehow. But it's part of an important trend.