Here is a chart for gain against angle of arrival for dipoles at various fractions of wavelengths high:http://www.karinya.net/g3txq/temp/angle_of_arrival_stats/aoa_dipole_elevations.png
For example on 20m, up to about 12 degrees you're around 5dB better.
However he has an advantage of being near the ocean for signals coming from Europe.
The things that come to mind are:
Filter insertion loss is 6dB so if he's using the SSB or CW filter and you aren't, he's 6dB down.
Difference in antenna height on 20m band is 6dB.
Difference in gain is ~4-5dB.
In short you've got around 16-17dB difference right there.
It could be that the proximity of the roof and the wiring and ducting within the loft is having a significant effect on the radiation pattern of his beam.
Then there's this from the FT1000 reflector:
"Typically I find S9-8 is 5 or 6dB then 8-7 will be 4dB or so then the rest will wonder around 2-3dB per S-unit."
I think there's your explanation. You could be both receiving a signal that would be S9 on your antenna but his could be 16dB down which could show a difference of four or five S points. Doing more reading it suggests that the S meters on the FT1000 below S9 are far from linear and far from consistently calibrated which is just another reason not to rely on them as an accurate indication.
And as a final point I would like to make from operating mobile, I can drive 5 miles down the road and have a S3 signal come up to a S7 and keep varying as I drive. Between my town and the next is what appears to be a sweet spot for 20m from the USA. Its not even on a hill but I can watch signals increase as I approach it and fall as I drive away.