Jillaine, I think you may be misreading me. I am not seeking a general change of approach. This is a fairly unusual set of circumstances.
The only direct record in this instance for LNAB is the baptismal record, which is clearly to be disregarded, and WikiTree guidance explicitly allows for that.
As I have said in previous comments, for less famous families, like my own forebears, I would seek to use the most common contemporary form of the name (though that is not always easy to determine - I have one 17th-century individual among my Cayley forebears who signed his name in at least 4 different spellings, a 2-page official record in which his name is spelt three different ways, and a burial record which gives yet another - and rather odd - spelling). I would draw on birth records where available, while recognising that parish register spellings can be peculiar.
Here we have a very prominent family where the baptism record gives a spelling it would be silly to use, where there appears to be a degree of variation in the contemporary spelling of the name, and where there is a norm used outside WikiTree, including in genealogical works - and which has been followed in all the other Wikitree profiles for the family apart from the children of Thomas who currently have the LNAB Cissell. My own reading of the WikiTree guidance is that it allows some leeway in these circumstances: you, I suspect, will disagree.
Shall we leave it at that? I do understand where you are coming from, but I am afraid we are not going to agree.