Polarized light from HD 189733 b? Maybe not?

Non-Detection Of Polarized, Scattered Light From The HD 189733b Hot Jupiter
Link

Abstract wrote:

Using the POLISH instrument, I am unable to reproduce the observations of Berdyugina et al. (2008), who claim to have detected polarized, scattered light off the atmosphere of the HD 189733b hot Jupiter. The amplitude of this signal is reported to be two parts in 10^4, which is larger than twice the amplitude expected for a perfectly diffusing, Lambertian planet. However, Lucas et al. (2008) find the polarimetric modulation of tau Boo to be inconsistent with a planetary origin, and any such planetary signal cannot be present at larger than 15% of the amplitude expected for a Lambertian sphere. I achieve photon shot noise limited, polarimetric precision of 2.4 x 10^(-5) each night, while the precision in eight-night phase bins from Berdyugina et al. (2008) is 6 x 10^(-5). Thus, my observations would have isolated the phase-locked component of the reported modulation with a signal to noise ratio of about eight. My non-detection shows that the polarimetric modulation observed by this group is not stable and cannot therefore be due to the planet. Starspots filling 1% of the stellar disk are known to exist on the active HD 189733 (Hebrard & Lecavelier des Etangs 2006, Croll et al. 2007, Pont et al. 2007, Winn et al. 2007), and they are most likely the cause of the signal detected by Berdyugina et al. (2008). Such planet-sized starspots viewed on the stellar limb may generate stellar polarization of order one part in 10^4 to 10^5 (Carciofi & Magalhaes 2005), which is comparable in amplitude and orbital phase to the reported modulation. Planet-induced starspots are observed on tau Boo in photometry with the MOST satellite (Walker et al. 2008), and there is some evidence that the planet orbiting HD 189733 also induces activity on its host star (Shkolnik et al. 2008).