Historical Problems of a Fantasist: N M Browne

My current work in progress is not historical but a kind of horror fantasy. While this might disqualify me from blogging about it on this forum, I would be a poor kind of fantasist if I couldn’t find a way round this minor disadvantage. I would like to claim that writing history and writing fantasy is very similar – no, damn it, I will claim that they are very similar and history and fantasy writers battle the same difficulties in bringing a story to the page and making it live in the mind of the reader.
But first I must acknowledge the biggest difference: research.
For most writers of all persuasion research is necessary – crime writers need to know about procedure and forensics and bus times and ballistics, even fantasy writers have to find out how the world of their imagining might fit together, which has had me researching dye making, wolves, insect life cycles, animal husbandry , beer making and physics theories in the past to name but a few. I usually ‘borrow’ ideas from history – Renaissance Florentine history for ‘Basilisk’, theories about Neolithic history for ’Story of Stone', maps of pre war London for ‘Shadow Web’ but I would hazard that historical writers probably do more research than most of us.
In writing fantasy after what is usually a relatively short period of research, I begin a more protracted period of invention ie I make it up. In writing fantasy the making things up only has to be constrained by what is feasible in the story world of your imagining, and credible to the reader. In writing historical novels there is that added constraint of all those by turns inconvenient, extraordinary and sometimes plain unbelievable historical facts: golden handcuffs manacling the writer to the ‘truth’ of the past.
Of course research is only the beginning. We have probably all read novels of one kind or another when the research gets in the way of the story; where at every point the reader is reminded in great detail of the thoroughness of a writer’s preparatory work but remains admiring yet unengaged. The same is true of fantasy where the reader is told about the world in such minute detail that they story is dwarfed by the intricacy of the world building. We have to make a choice not of how much but also of what to tell and that really is the hard part. For me the trick of fantasy and of history is to introduce the reader to an alternate reality - the past or another world, as if it were the only reality – immersing them in strangeness without explanation or exegesis. In this kind of writing each carefully selected detail of clothes, furniture, manners builds on this picture of otherness.
Which brings me to the battle for the reader’s cooperation, their willing suspension of disbelief. How far should you go? In conveying the truly strange where do we strike a balance between what was or would be and what will be meaningful for the contemporary reader?
AS you may by now have guessed I am struggling with this problem in my current project. In constructing a reality, you can’t easily abandon its logic for that logic is the fantasy version of a historical truth. How strange can the story reality be before I lose my reader? So I would like to ask you historical writers how far can the illusion of history found in a novel truly reflect the alien nature of other times and other societies? Should a research truth be abandoned in order to tell a better story. In order to be credible and feasible do we have to ignore the most peculiar of facts? How do you deal with the truly weird?