I like it. There is really no reason why one couldn't add as many details about a word as they see fit.

Here is a sample of it with a noun (BTW, +Matt Dinse , you should make those alternative entries you put in there)
http://middangeard.org.uk/vqp/doku.php/ns:bainas

Devil's advocate: I'm a bit reluctant with regard to adding mutations for verbs with i preceding, as we have few attested examples. The standard assumption would be soft for singular, nasal for plural, and indeed we have Gyrth i chuinar in 1972 ... but Gwerth i Guinar in the early-50s Grey Annals. Other early-50s texts like the King's Letter are often taken as granted for Sindarin canon, but I hesitate to standardize i-guina, i-chuinar as Tolkien may have wavered throughout various forms throughout the decades. mín i gohenam and ai gerir (VT44) are the only other examples I recall. Though perhaps I'm being somewhat influenced by Salo's neo-S calad ammen i reniar instead of idh reniar ...

I'm not against labeling verb classes (though different people probably classify them into different categories, or use different terminology: so consistency might become a problem), though I wouldn't go so far as to add tense conjugations. People subscribe to different theories and some will say *cenn for pa.t. of cen-, others *egin, for example. It would certainly be useful to distinguish classes as those distinctions can affect the conjugations of homonyms, e.g. Q from -tă, -tā, and causative -tā.

I've been adding "i/in" to the verbs not to put my foot down and say, "this is how we'll do relative clauses from now on!" but to illustrate the two most common types of mutation. It's convenient, I think.

I-verb and A-verb are pretty universal classifications of Sindarin verbs, as far as I know. Getting into more detail than that is where people have their own little theories.

I agree about the past-tense for I-verbs, that's still murky right now. But, we could make lists of "past tense conjugations that you may see people using" for them - that may be helpful.

Oh ok, that i/in makes sense.

How would you classify gala-? It would be an "A-verb" but is conjugated  galon, gala / gâl with a strong pa.t. aul/angol instead of *galant/s (I'm not sure if it's transitive or not). Though perhaps "A-verb" doesn't directly mean "this is a derived verb" but rather "this doesn't form the present tense with -in, -ir (etc.)" ? I must admit I'm rusty on Sindarin terminology.

Arguably, though, I've been told I'm quite good at finding reasons why something might not work, but not so good at suggesting solutions. >_<

There can be A-verbs and I-verbs which act like each other due to analogical changes, analogical "errors" you can play with to flesh out a character's history and tutelage in Sindarin. But, that's not really part of lexicon, so I don't think it'd really come up. Especially since we are coining the terms, so we can assign variations and quirks to the verbs as we like.

I would only keep information which is absolutely necessary, namely special case mutations, peculiar plurals (see buðu etc.) and unpredictable past tenses (e.g. different augments: nor-, onur, but tol-, *udul).
In the long run, it will save a lot of work: As soon as new information on pronunciation, mutations or plurals will be published (as I'm sure it will be!), one won't have to go through the whole dictionary to make the changes.

The classification into a-verbs and root verbs refers only to the present (I'm actually not entirely certain where the causatives belong), in the past there can be mixed conjugations. But it's completely redundant to mark it, as it can be read off from the stem.