Your analogy is wrong, more evidence just whittles down the suspect list. If all you had was a partial print, you'd have alot more suspects than if you had a thumbprint and DNA evidence. Which is much the same with the weather, data is essential, if something is missing data than you naturally have to be wary of it's results especially when everything is unpredictable. If it manages to luck into being right with less data, it does not make the idea that less data is more right, it just means that the model was right that time.
With the models being unable to pinpoint anything right now, surely you'd be more confident if a model had more data being fed into it? Rather than confident of something that has less. I just cannot see why anyone would put more stock into the 06z/18z at this very time unless the results it spews out aligns with your own weather bias.