I'm just going to venture a standard guess and say your laptop is a fixed resolution so in order to display something the same size, windows is scaling it. In order to really compare resolution by itself you need to remove scaling. At the very least it certainly isn't going to be done on the same display since they can't change their native resolution so scaling will always be involved.
The best way around this is the detailed test I described above where the projectors differed very little and the scaler that was used was high end. Actually, that test was more unfair to the 720p projector since they were both fed 1080i (1080p native proj had only to deinterlace whereas the 720p had to deinterlace and scale). That's what makes it so compelling in that people had a very hard time differentiating on a giant screen and if anything the 720p was shortchanged in the process.
I'll also add that there is a lot of variance in methods of scaling. Not all scalers are created equal by a long shot. The ones in most displays are very poor (i.e. this is why many $100-200 upscaling DVD units outperform most displays) as is the way windows handles it although many people will used 3rd party apps to arrive at better results.
Panasonic 65" VT60 (calibrated by Jeff Meier), Marantz 6005, Kef Q towers/rears Kef Reference center, Win 7 Pro with XBMC, Harmony Ultimate One
Xbox Live: Arbitrage XXL