Quote from TardisGreen >>
So with Glimmerroot you get a random card from your opponent's starting deck (assuming you pick the right one) and learn nothing about their current hand.
But with Envoy, your justification for not using it is that your opponent's cards won't be useful anyway?? So what good is Glimmerroot then? Envoy actually tells you what cards your opponent is and isn't holding.
OP asked about Envoy and people's experience with it, not Glimmerroot. Some of you sound like you're just commenting on a card reveal thread and you've not even tried Envoy. You're so self-assured you're comparing it unfavourably to a worse priest minion from 2 years ago that you probably never even played.
Assuming you knew their deck archetype (and you always knew that in open deck formats), Glimmeroot gave you a random card 100% of the time. Envoy's chance of getting a card are always less than 100%.
The info you gain with Envoy is largely useless. Good players are already making reads on what cards are in or not in their opponent's hand; they don't need this extra info. Bad players aren't going to be able to utilize the info anyway; they can rarely make the best decisions even if their opponent's hands were completely face up, usually because they are playing with their "creative" home-brew decks that are extremely underpowered to begin with. And if this kind of info was useful, why has Chameleos NEVER seen competitive play?
Finally, but most importantly, what archetype would utilize Envoy? Clearly not Rez Priest nor Combo Priest. Is there some Unicorn out there? I suspect not. Even if Envoy were a "good" card, it has no deck to go into.
First of all Glimmerroot was not always a 100% chance to get a card, as several classes had multiple archetypes while it was in rotation (for example Mage which had Quest, Control, Odd and Tempo variations all at once). Not to mention times in between metas when people were experimenting with all sorts of stuff.
Second, there are countless situations where Envoy will have a 100% to give you a card. The most obvious example being if it shows you The Coin in your opponent's hand. This is where the thinking comes into play, which people who completely dismiss the card seem to be missing.
Thirdly, knowing what cards your opponent has or doesn't have is BETTER than just assuming what they might have. It may not be enough to make a card competitive, but this is still just simple logic that anyone should know.
And finally, you can use Envoy in almost any priest archetype because it is a tech card. Not an archetypal one. But atm it is most useful in Galakrond Priest since it can snag you additional Galakrond cards, cheapen your Mountain Giants if you're playing them, and potentially add minions to be summoned from your hand by Princess Talanji.
I was often able to determine if my Mage opponent was an Odd Mage before the first card was played. There is a little trick you might not be aware of. Quest was usually determinable quite early on as well. What's left? A Tempo Mage that whiffed on turns 1 and 2 and looks like a Big Spell Mage? Even then the pool of cards offered might give the deck type away.
These situations where your opponent hasn't used their coin on turn 1, how often do they come up? How often will the Coin be offered as a choice? (Can't be very many - opponent probably has four cards in their hand.). And how valuable is a 2/2 that draws a Coin? Is it game winning? Don't think so.
More generally, even if you guess/infer correctly, how often is the card really useful? If you are up against a Face Hunter, is playing a 2/2 on turn 2 and getting a Leper Gnome gonna help you survive?
And last, but not least, Galakrond Priest is a meme deck. I realize the Hearthpwn community is full of "creative" types that love to swim upstream. But Hearthstone is a game with a winner and a loser. And if you are playing Galakrond Priest, you are playing to lose.