Microlensing results: Jupiters common?

Saw this article linked on EPE, unfortunately it is behind the Nature paywall so I don't know much about what is going on here.

Seems like it might be quite interesting though...

The article point out that out of 13 events of a star eclipsing another star and producing microlensing that were recorded between 2005 and 2008 five resulted in planet detection with one of then having 2 planets.
Micolensing is sensitive to planets that are located at 2.5AU, for star that have a mass 1/2 of the sun.
The author argued that even with this limitation more than 33% had planet and that is a fraction of the planets. This shows a larger probability than the one returned by RV measurement
The planet have masses between 50 and 300 time earth.
The planets are probably locatd behind the snow line, where they were formed.
This later suggest that either: most planets do not migrate very far or that the most migrating planet get destroyed.
Microlensing is sensitive to Jupiter size planet and none were found, so it points that the mass of newly form gas giant is skewed toward low mass object. One of the planet is 13 earth masses. This skew is in contradiction with the gas giant formation theory

The article is 1 page and does not mention wich stars, prossibly just statistic from the planet encyclopedia

Thanks Roland. Very similar results to another paper, also by microlensing.
http://arxiv.org/abs/1001.0572
In this article, the authors analyzed 15 micro-lens events (2005-2008) and found among these two systems with gas giants (one of them even once with an analog of Jupiter and Saturn).
However, the additional sensitivity analysis, the observation of micro-lens events, the authors found that the frequency of gas giant planets in wide orbits may be around 30% only around red dwarfs.

On the other hand, these results are similar to the figures of the future work of the AAT survey
https://solar-flux.forumotion.com/extrasolar-planetology-f5/frequency-of-jupiter-analogues-t714.htm

Quote :

- 120 non-detections implies upper limit of 37.2% for planets with K > 10 m/s in 3-6 AU

But it's not for red dwarfs, and yellow dwarf. Although of course the radial-velocity method is less sensitive than the method of microlensing.

In addition - John Chambers, not a specialist in the region of microlensing. Most of his articles on the search for RV-planets on the Magellan Telescope.

http://arxiv.org/abs/1001.0572
The authors of this article, really specialize in micro-lens survey.

That is the base article for the Nature news presented byJohn Chambers