On this one (on the old site) I followed a standard approach I used of putting my guess at species level ID (first) and then the safer (generally higher taxonomic grouping) ID. On Old iSpot this was working with this entry having a LikelyID of ‘Moths (Lepidoptera)’ .
Possibly one of two things have happened here (or both). It could be that new iSpot takes the first ID a person suggests as their suggestion and privileges that for LikelyID in the absence of agreements. Alternatively it looks like the second ID is being treated as being in the catch all ‘other organisms’ category rather than in invertebrates where it should be. Since it is not possible to have ID on ‘other organisms’ that means that ID is losing the fight for LikelyID.
If the second it is not clear what the bug is. It could be that the generic ‘Moths (Lepidoptera)’ is wrongly specified in the dictionary (as other organisms) or it could be some fault where a second ID by the same person is being categorised as other organism.
On old iSpot it was not possible to have reputations from different groups applied in one LikelyID calculation - essentially it used you reputation in whatever group the poster had assigned their entry to (Invertebrates in this case) so even if on old iSpot ‘Moths (Lepidoptera)’ was wrongly classified in the dictionary it would have worked in this entry.
 Note at the time of writing the only IDs were my two neither of which had attracted agreements. Once people start agreeing it might well change.
 by design although I’ve never been 100% convinced that is the right design choice