Re the two unknown to you, it often pays in the early stages of such research to include some "in-betweens" who may or may not have a connection, whether directly in the family line, as it occasionally helps when "brick walls" are encountered in the main line.
Also particularly with census information, it is less formal and less rigorous generally, so misunderstndings and errors can creep in, at the census time, compared with the legal implications of registering Births, Marriages and Deaths.
Additionally, there can then be mis-interpretations and mistakes in the modern transcriptions from any of those (ancient) documents in to the computer databases making up the various Search Indexes.
I have seen a large range of errors, so better to include a little mtoo much generally, than to ignore them.
I will try later todayto harden up on Alexander Crighton, to see if I can make a more solid connection in to the Scottish Records.
Although there are Crightons on Castle Garden, the predecessor of Ellis Island, at present none seem to be him.
www.castlegarden.org (19th and early 20th century)
www.ellisislandrecords.org (late 19th century to middle 20th century)