The fact that there are children named who you don't know about isn't surprising given they were non-conformists and there are no baptisms. I've seen many wills where not all the grandchildren get a bequest, so I wouldn't read anything into that. I don't find it odd at all - perhaps Mary was his favourite. Were there any great grandchildren in 1839? Again it wasn't unusual for any great or great great children not to be mentioned.
If John specifically mentions granddaughter Mary Sharman then there's every chance that she was the daughter of William and Mary, and thus Mary (Gurry) Sharman was John's daughter. Slightly odd that his daughter Mary doesn't get a mention - perhaps she died before 1839 and her eldest daughter got her share, although I can't see a burial in Beds. In fact I think this is the most likely explanation - that his estate was divided up between his children, and the child of his deceased daughter got what would have been her mother's share. Susan and Sarah both appeared to have married in Lt Staughton before the will was made - were their married names mentioned in the will, or were they just referred to as my daughters Susan Sarah and Ann?
William needn't have died - he could merely have declined to act as executor. I don't understand though how John acted as executor - the May attachment may have been the Grant of Probate.
William aged 60 born Lt Staughton was living in Eynesbury in 1851 with wife Hannah.
It looks to me as though John born c1813 at Lt Staughton was the son of John age 57 living in Eaton Socon in 1851 who in turn was the son of John whose will you have. The difficulty is in proving that the youngest John was the son of John born c1794.
There must almost be the possibility that John born 1794 wasn't the son of John and Sarah, and that John born c1813 was their son - born c25 years after they married so not out of the question. I'm not suggesting this was the case, as I don't think it was, but it can't be ruled out.
David