A website at Duke University has all the originals of Charles Wesley's
published and
manuscript poetry, with annotated transcriptions that are searchable! A short history of each publication is there too. Most annotations show the changes the Wesleys made. If you put all the transcribed publications in a folder and use a good search program (like
Agent Ransack), you can find
anything.
About archaic language, I personally feel that the archaic language that is commonly understood should be kept when reporting the original text; words not in common use today might be substituted with footnotes. For church use, I prefer to substitute most archaic uses unless there is a teachable reason ("Ebenezer" for example) – but that will vary considerably depending on history and tastes of the congregation. To me, the "fake archaic" language of some 19th-century hymns borders on offensive, but might be helpful to others.
Contractions are a different matter; I would retain only those few commonly understood (or used elsewhere), such as "e'er". So unmov'd -> unmoved, prest -> pressed, etc. This is largely because some search engines won't translate, or do it poorly — most don't find things like mov'd or disus'd. Of course, when laying text under music, more contractions are often used to conserve space; and some people prefer to use contractions needed to reflect the meter (I don't).
Barry Johnston
Gunnison, Colorado