As a frontier post, it might be possible there'd be better defences around the doorway externally than there are currently. Indeed, it's possible there'd be no ground-level access at all, with the main doorway on the level above that, accessed by a narrow stairway externally. Certainly, reworking the building as an inn might alter the access options, but some of the defensive structures would likely still remain. That'd be a lot of extra work here though, I realise!

There is a problem with the fireplace walls defensively now though, in that these are both thinner than anywhere else, and heavily advertised on the outside by projecting beyond the main line of the walls. In reality, the fireplace flues would likely have been built into the walls to disguise this (and probably be smaller than are currently shown), or be reinforced externally to have the same thickness as the main walls. Orcs and ogres might be reckless, but they're also not necessarily as stupid as systems like D&D have sometimes portrayed them 😉

The vertical striping could be due to the shape of the roof, with the underlying rafters pushing out those parts directly beneath them, so leaving a hollow away from that (the difference is likely to be very small, so wouldn't show as variations in the roof's appearance otherwise, except maybe under very low-angle lighting). I agree the rafters will be likely the defining point though.

Keeping the look fairly straightforward sounds like a good plan!

Having lived only in places with tile or slate roofs, I can confirm Shessar's photos are indeed entirely accurate for both those materials. Slide, wind effects and thin ice/snow melt basically works from the top down, and outer roof edges facing more nearly into the current wind, so those will all tend to clear of snow first. This can be enhanced around objects sticking out of the roof like chimneys and stove pipes, especially where those are in use. There may also be some smoke discoloration of the snow near chimneys that are in heavy, regular use as well (albeit that also tends to mean snow there will melt faster as well - introducing foreign particulates to the ice/snow helps it melt faster generally, like applying salt to road and path ice).

Roof patches do tend to be harder edged than you've illustrated so far, and with a tendency to remain in the hollows a little longer than the ridges on shaped tiles, as you've already noted. There are a lot of variables however, and commonly, once the snow's started to melt on such roofs, it will tend to clear fairly quickly thereafter, unless there's fresh snowfall heavy enough to fill-in the cleared gaps.

There are things to be said for both options (plus of course, you can't assume everyone will wish to map somewhere where the snow falls in late autumn/early winter, thickens and only starts to thaw around the start of spring - i.e. the randomness of what happens in British winters!), and both are obviously buildings. However, the one with less snow cover has a bit more character to it, which would definitely support the idea you mentioned in your most recent post Sue.

Yeah, I'd settle on one or the other, I think.