Well, that was maybe naughty of her, but she may have misunderstood you..
Delayed echolalia is, well, tricky. It's not strictly speaking typical but I hear anecdotally that there is more and more of it being seen in clinic. In my moments of random musing, I do wonder whether the experiences of young children have changed in a lot of ways and there is more scope for it to develop. Every time we see a digger in this house I launch into: "diggers were made for dig, dig, digging, scooping up the earth and lifting and tipping, they make big holes with their dig, dig, digging, they can work all day"... ds is only 18 months but already jargons this sometimes. In 2-3 years time, he may very well be able to "parrot" it, but it would not necessarily be a problem iyswim?
I am not saying that your dd has delayed echolalia of this type.. but I think it has become harder to just see echolalia clearly as a red flag. Things like "Dear Zoo", the "Gruffalo", certain phrases from certain television programmes (with American accent! Ben10, Dora etc) seem to be part of the "repertoire" of many preschoolers now. I used to automatically raise my eyebrows if a child demonstrated this type of linguistic behaviour but more and more, there are children who are presenting with this.
There is a point in language development where formulaic language use is quite typical. I think the easiest example is "whatdat?" for "what's that?" or "uh oh" or a half-said/half-sung nursery rhyme like "twinkle twinkle". Snatches of language used in context but which the child doesn't really understand but uses communicatively? For a textbook child, this is confined to the very early stages of language development.. e.g. the beginning stages. For some children, this formulaic stage lasts longer and for some, it can become a crutch. For others, it may be that they just process and learn language in a different way to that "textbook" child - they are processing language "top down" vs "bottom up", or are a "gestalt" language learner. Many people who are on the autistic spectrum process language differently, for example, and many are "top down" processors.
However, it is worth saying here, I guess, that what I guess we have to remember is that there is not a very substantial body of literature on individual difference in language development. We know a bit about how language typically develops (and this is pretty consistent across languages and cultures) but not enough, and what we definitely don't know enough about is those children who appear to be developing atypically who spontaneously resolve. I have seen some kids in clinic who appear to have absolutely DIRE language/speech/communication at 2.5 or 3.5 who have no outstanding issues at 5.. and these are the ones that actually make it to clinic. Many don't.
Generally speaking, delayed echolalia that makes sense/is in context/is appropriate to the communicative function is really not as much of a worry as the type of self-stimulatory echolalia that the phrase "delayed echolalia" brings to mind e.g. just randomly reciting a television script out of context, at least not in the preschool age. However, it is not typical and is something that needs to be monitored/kept an eye on. To coin a phrase, it may be something or nothing, and it may be too early to tell.
In the preschool years, subtle difficulties are very, very difficult to pin down. And there is cautiousness because of the variability in how it can progress, particularly in very verbal kids.
What I guess you need is support and help on how to deal with the behaviour and develop her language now, right? I don't work with this age range so much anymore so I am not so "up" on what's out there (I work mainly with secondary kids). What are your main concerns about her language?