I think that's a great point about working towards a minimal viable product. I was shocked to read recently about all the wasted time spent doing things like implementing versions of Python and Javascript on Parrot instead of working to make happen its whole reason for existence, Perl 6.
I've just read that article and I feel there's a subtle re-writing of history going on there. I stopped following Parrot shortly after Dan Sugalski left (somewhere around 2005?).
I wasn't there before the original Parrot announcement but I did start lurking on the mailing list around 2001-2002.
I recall Parrot always being touted as an open VM for all dynamic languages (this is before dynamic languages started to appear on the Java VM). You have to put it in the context of the rumoured .Net VM that Microsoft was about to bring out—(and then did shortly after the Parrot project started to gain traction)—those of us interested in open software were keen to have our own.
Sadly Parrot stumbled and the rest is history. Mono wrote the obituary.
• The Perl 6 spec was still being argued about (plus I recall Larry had been seriously unwell).
• The intention was to create an open VM for all dynamic languages.
• Perl 5, Python, Ruby and Javascript were always targets for Parrot.
My frustration was that, instead of putting everything into creating usable Perl 5, Python and Ruby implementations (Javascript was pretty-much browser only at the time) the Parrot team faffed about with toy language after toy language.
I may have imagined this as I can't find reference to it online.
http://whiteknight.github.com/2011/09/10/dust_settles