Don't know if you've encountered this resource, but I found it quite helpful for inflections, and paradigms in general
http://faculty.virginia.edu/OldEnglish/courses/handouts/magic.pdfI think the "-anne/-enne" variation is related to particular dialects/times/personal preference; but only because I've not yet noticed any systematic variation in it from reading OE, and none of the modern books on OE I've read have articulated a rule! Some of the other variations you've noted may have a similar explanation, although I can't immediately recall having seen "to" before a verb without the verb taking one of these endings, or "-nne", in the case of the contracted forms with infinitives ending "-on", although it does ring a bell. I'd be interested in a specific example. The inflected infinite has specific uses, so maybe in the exceptions, the "to + verb" combination is being used with a different meaning?
Something else I always keep in mind when, say, there are just one or two recorded exceptions to an "established" rule, is the assumption that when OE was written down, no-one ever made any grammatical mistakes (we know scribes did, occasionally) and that later copies were always verbatim copies of flawless originals (we know they aren't).