2 is as far as I can tell anecdotal and subjective. There is after all no evidence that there is no evidence of interest in roguelikes that are not heavily combat oriented. .... and the avoidance of combat cannot exist without a reliance on combat.
This is a question of your standard of evidence. After all, there is no shortage of attempts at games, mostly so-called 7DRLs, that purport ..
I'll interject at this point and note that "attempts at games" is apples and oranges to "actual games". As you suggest later, a 7DRL can only provide a casual and shallow gameplay experience at best, due to the limited time available. If they do demonstrate alternatives, it proves nothing other than the existence of a fun casual gameplay well designed.
.. to be roguelikes and try to make things less combat oriented. As far as I can tell, none that fit the bill seem to have gained a nontrivial audience (say, on par with a top five angband variant) or spawned successor projects that have either. Maybe this says more about 7DRLs than anything else, you might argue. But it's also true that, as I mention above, Sil has been praised for its progress in the areas outlined by Omnivore. Yet Sil, as I understand it, cannot be won purely by slinking around avoiding combat and no variant has come along to challenge that situation, again calling the demand for less hack and slash roguelikes into question. The fact of the matter is that there's been plenty of opportunity for such a game to emerge and nothing seems to have happened. This is a reflection of demand.
No, your last sentence is arbitrary and not supported by your preceding argument. I argue that it is a reflection on availability.
You can look at all the existing RPGs and roguelikes out there and even wider out into FPS and see that for all the worthwhile ones like Oblivion, Dragon Age, Baldur's Gate, etc etc combat is the crux. It gives the meaning, the focus and is generally a shallow and achievable core mechanic. And it is doable because it has been seen to be done. Someone can just clone it and do minor iteration on it, and it is conceivable in their mind how to create it from scratch. Conversely, there are no equivalent non-combat reliant games to clone and iterate. Making a game from scratch is a tremendous investment of time and energy, but coming up with a new genre that is a superset rather than a subset of the combat-reliant standard, I argue is inconceivable to most. If someone made it, and it was done well, it would be a breath of fresh air and in demand.
In order to support an argument of lack of demand, supply of quality goods needs to be disdained by the target market.
re: no avoidance of combat without combat, this seems to be the real issue. If there is no combat, it's not a roguelike anymore. If there is combat, the conventions of the genre lean heavily toward engaging and winning at it, not avoiding it. One radical approach would be to make it impossible to win in combat in the long run so that alternatives would be unavoidable, but the stable of models for this kind of thing is pretty thin. Stealth and pacification (e.g. crawl's Elyvilon) seem to be the only reasonably developed alternatives and from what I've seen only the stealth option seems very compelling.
I agree. Combat is essential to a roguelike. Even if one doesn't need to engage in combat, for the standard game world, it's presence lends a feeling of believability. Stealth and pacification are merely decoration on the combat cake however. Combat as an option gives value to other non-combat related choices that may rely on the inputs and outputs of others doing combat.