Dear Ravi,
The reason why linguists in general are
dismissive of Paul Manansala's conceptions has nothing to do with the "Aryan"
problem. Most linguists don't give a damn about the "AIT/OIT" shouting match and
certainly wouldn't blame anyone for not taking sides in it. The actual
reason is the same that makes archaeologists ignore Erich von Däniken. To put it
bluntly, Mansala's "linguistics" is rubbish. Not because it's done by an amateur
-- the methods of historical linguistics are not so esoteric that an amateur
should be unable to understand them -- but because he uses no rigorous
methodology whatsoever.
Austronesian is safely the largest bona
fide language family on this planet (I don't count Niger-Congo, which is not
even nearly as well-defined) -- it has about 800-1200 members (depending on the
criteria used). Its time-depth is comparable to that of Indo-European, its
internal taxonomy is highly complex, and some of its branches have experienced
very thoroughgoing sound changes. It is simply not enough to list ten or even
twenty similar-sounding words from arbitrarily selected modern languages in
order to establish an Austronesian etymon, as if comparative analysis had
never been invented.
Manansala is also blind to the difference
between similarities due to areal effects and to substrate influence, and
systematic correspondences due to common descent. This visual handicap is what
allows him to treat Sanskrit as a Dravidian language. No real historical
linguist (and no informed amateur worth the adjective) would confuse such
things.
Internet resources are a mixed blessing.
You can find trash as well as reliable information there -- the former more
often, needless to say. Things that would never get published in a respectable
journal can be circulated easily and quoted as "information"
Piotr