These are both common problems people have as they start with game design.
1) You will think about things like "there will be a spooky corridor here" but later as you try to make it you notice you don't know what to put there so that it would be "spooky". As you gain more experience in game design you will learn to visualize and think out the different assets and features in the extent that you will actually need them. Improving on this translates into trial and error pretty much although part of it is taking influence from existing games as you play them. You just go on making levels and props and you'll widen your toolset as you go.
Thats right this is related to what Im saying. It seem that you
have to walk a long "path of trial or error" probably, to be able to do it.
Lets say you a coder -
if you do not cross (i mean something like travelled, sorry for weak english) that path you will not do it , and if you travelled it you will do it Lets say you are a designer if you do not cross that path your advices are useless, so it is some kind of schema
notpath coder notpath designer
path coder path designer
some combinations are bad,
I am thinking also "if the trail and error path" is necessary for coder as a indyvidual experience then some external separate
designing my be totally not usefull - but those are maybe not to much clear things - but i wanted to state of crusial importance of indyvidual experience which my make the advices of separate designers quite unusefull (useless).
(now i am somewhat lost in what i was trying to say and what i was trying to ask :/ but there is some topic here )