A lot of people are busy mining sentences. However a lot of the time there are two versions of the sentence a sentence with as many kanji as possible, plus a kana version. However often there is no version of the sentence exactly the same as what is in the book. Now I have no problem with there being three versions of the sentence, but I think not having a version of the sentence exactly the same as what is in the book could be a problem due to several reasons.
1. It is more difficult to match the sentences up.
2. It is more difficult to check that the sentence has been correctly entered (especially if you want to check a large number of sentences).
3. The sentence might be over kanjified. ie the sentence might be technically correct, but a native speaker might not write it that way.
4. Most people entering the sentences are still learners, they could enter the wrong kanji causing a lot of potential confusion.
5. Too many kanji might make the sentence more confusing/difficult for those who like to balance kanji study with study of grammar.
I am interested in hearing other peoples opinions on this matter? Of course I expect some people will say if you don't like the approach, then enter your own damn sentences !! Anyway this isn't a complaint just an opinion and I hope other people share their thoughts on this.

Isn't the whole point of the so-called "sentence method" to keep you reading and maybe writing correct, meaningful Japanese in a controlled environment? If that's the case, I don't see how producing you own sentences could be the right way to go... The way I understand it, you should be mining sentences from reliable resources, such as good textbooks, dictionaries, grammar books, and so on. I don't think you should try to "kanjify" the sentence though. Even though we feel really confident in our kanji abilities after RTK1, if we kanjify it, we might run into problems number 3, 4 and 5 that you enumerated. So, I think it should go like this: Question = original sentence; Answer = kana sentence (readings), and the optional fields (use those you think are necessary): meaning, definitions of words, notes, etc.
Of course, none of this solves the problem of sentences you mine through audio resources, such as doramas, animes, podcasts... But I think that beginners should try to stick to written sources for sentence mining at least until they have been exposed enough to the language to "get a feel" for how it all works.

yeah, if I don't see a 'uk' tagged on, i'll prefer seeing the kanji because even if some people don't use the kanji, there are those who do. Although I have seen people use kanji even though it was 'uk' like 既に which I didn't catch until I used rikai chan

My only argument for "over kanjifying" assuming it's the correct kanji, is that part of our studying via the sentence method is getting the reading of the various kanji. The more you pack in a sentence, you more you learn.

As for this creating a bad habit, I don't think so. You SHOULD be getting 10x or more as much input from other good sources. This input should over-ride any minor bad habits you picked up by using too many kanji (assuming that's a bad habit). For those that are UK.

As for the topic, it is best to follow the book IF we're doing group efforts. Use the kanji, kana and translation as provided. For your own, hey, it's your training and not ours.

I think problem 4 is solved by simply typing the kana into a dictionary and making sure the correct kanji comes up ..i do it at work all the time when I am typing out emails in japanese and i hit spacebar and some crazy stuff pops up(IME)

We shouldn't be training for amazing kanji recall, we should be training for amazing Japanese. The best way to do this is to use Japanese as the Japanese do. We don't want our written sentences to look unnatural. I think even Khatzu was caught out with using too much kanji when writing. Learning kanji is a step in the method, not the goal. Authentic sentences from authentic sources are the key to this goal. Don't over kanjify, it looks unnatural and won't help recall that much.