Quality of current OLPC Software

There seem to be many conflicting "fact" opinions around this article, 
which seems not constructive, and the personal comment about 
Negroponte appears counter-productive.  It is almost like BP has taken 
lessons out of the old big bad IBM FUD book.

Previously, those writing about using the OLPC have seemed enthusiastic 
and happy with it (eg BBC Nigeria coverage 
http://news.bbc.co.uk/2/hi/technology/7115348.stm ).  Now following the BP 
punditry link, some of the technical comments (mainly N Weaver) are very 
negative about the Sugar & Linux implementation in contrast to previous 
positive news coverage.

So are review articles really failing to notice, that you can't save files  
or install applications?  How would you keep the notes in class mentioned 
by BBC?  PJ's take 
http://www.groklaw.net/article.php?story=20071223132431291 
http://www.groklaw.net/article.php?story=20080107182525297
Michael Tiemann's Blog http://blogs.cnet.com/8301-13507_1-9837995-18.html 
surely someone who created GNU C++ is going to notice such defects.

With Asus EEE PC, Intel Classmate, Shuttle and other low cost PCs being 
produced; is it really so surprising that M$ think  
they better have an OS available on low cost hardware?  Can a hardware 
project realistically tell M$ "No you cannot port XP to the XO!".  This 
insider Blog points out it is M$ doing the port, not OLPC 
http://blogs.technet.com/jamesu/archive/2007/12/05/olpc-i...  

So there's some supporting evidence for Ivan Krstic's rebuttal of BP's 
article "The paradox of choice" 
http://radian.org/notebook/paradox-of-choice

OLPC http://laptop.org/en/laptop/ explains the project has a specific 
commitment to FOSS software, so "cooperation to ensure linux image is 
easily re-installable" rings true to me.
 
Is it right to presume, that innovative software shipped today has no 
chance against something presently undeliverable (until 2nd half 2008 
http://www.news.com/beyond-binary/8301-13860_3-9829735-56... ), which is 
not designed to support the constructionist educational aims, and probably     
needs extensive pirated software to keep the end users happy?

If the FOSS community reacts with fear and hysteria, and apparently  
concedes defeat so easily; does it look like real confidence in it's 
product to outsiders? And are we really suprised when commercial companies 
distort the truth "wave of SCO-like problems"?  Time may prove Bruce 
Perens right, but the message appears too defeatist and likely to cause 
dissension, rather than be a useful warning to non-Free dangers.  If 
attacks on Negroponte deter FOSS community support for OLPC then 
http://laptop.org/en/laptop/software/ cannot succeed, and BP's prophecy 
becomes self-fulfilling.