I've had some success converting lens distortion parameters from Adobe format
to Lensfun format.

The main complication is the difference in coordinate system. This is described
in the Adobe lens model doco, and I go through the details in this thread:
http://forums.adobe.com/message/3382077

Here's how the conversion looks for the 2-parameter Adobe model, which maps
nicely onto the lensfun poly5 model (assuming a 4:3 aspect ratio):

k1, k2 and FocalLength (aka stCamera:FocalLengthX) all come from the LCP file
k1', k2' are the lensfun poly5 parameters

amax = 0.5 * (1 / FocalLength)
atol = (4/3)/(amax * (1 + k1 * amax ^ 2 + k2 * amax ^ 4))
k1' = k1 / (atol*atol)
k2' = k2 / (atol^4)

[here's some R code that calculates my example:
http://ouabain.org/lensfun/fit-adobe-lensfun.R ]

An example LCP file: http://ouabain.org/lensfun/DMC-GF1.lcp
The converted lensfun parameters:
http://ouabain.org/lensfun/slr-panasonic-local.xml

I compared the results of the two profiles (in Adobe Lightroom and rawstudio
for lensfun), and they match well:

Lightroom: http://ouabain.org/lensfun/p1010547-lr.jpg
Lensfun: http://ouabain.org/lensfun/p1010547-lensfun.jpg
difference: http://ouabain.org/lensfun/p1010547-diff.jpg

[n.b. to calculate the difference, I first scaled down the Lensfun output, because
there is a difference in the final scale/crop between Lightroom and rawstudio]

Original raw file here: http://ouabain.org/lensfun/p1010547.rw2.bz2

[for the Lightroom conversion, I first used dcraw to generate a 16-bit linear
TIFF, rather than let Lightroom convert the RW2 file directly -- otherwise Lightroom
will apply the 'standard' correction for this micro-4/3 lens]

I'll have a look at the chromatic aberration and vignetting parameters next,
and if that goes well try to write a generic LCP to lensfun converter.