elephant gambit or Latvian

The Elephant seems to be under a cloud at the moment despite the efforts of the authors of "The Exhilarating Elephant Gambit". Not only is there a line for White after 1. e4, e5; 2. Nf3, d5; 3. Nxe5 in which Black seems to end up with a difficult slightly-worse endgame (maybe the nineteenth-century 3..., Qe7 hangs on, but very few people play a gambit to snatch material then defend against an attack) and there's also this to contend with:

which is recommended in "The Mammoth Book of Chess" so it's likely quite well-known.

There's a really good recent strand on the Latvian by @jcidus on this site's opening forum - I'll find it and tag you.

The Latvian is literally losing by force, so there is not much of a choice here.

The Elephant is not exactly the most solid gambit out there, but it is playable after either 3.exd5 e4! or 3.Nxe5 Bd6 4.d4 dxe4.

Pure prejudice, my friend. The Latvian Gambit is a lethal weapon if you study it deeply, it might be the opening that gives you the best results.

Forget what the engine says; no one in all levels knows how to refute it in practice, and it's extremely annoying to face with White so much so that many players opt for the d4 line that leads to an easy equal game for Black.

The power of the Latvian Gambit lies in the fact that Black takes the initiative from the very start. Even being down material, Black gets a highly dynamic position after the move f5, allowing them to play for a win from move one.

elephant gambit reminds me a bit of a worse off spielmann. With best play , it seems black has just enough compensation for the pawn to not lose by force and at least his pieces look healthy but it is not pleasant.

latvian like many other early f5 gambits in double KP openings is just a relic of the old days. IT was never in sound footing but cloud engines buried it for good. 
At least if lord knows why you feel compelled to become an authority on the elephant gambit and learn lines inside and out that have never been played before as well the critical lines, you may get some reasonable success as well as the occasional surprise factor. The latvian is 100% surprise factor.Once that wears off and your opponent actually bothers to memorize a refutation line, you can never play it agaisnt him again.

Negative. I used to always play the Latvian Gambit in classical time controls in 2010-2015 , and my opponents would study me. Some of them took me all the way to move 18 of the theoretical refutation, and still lost the game because I had knowledge beyond that point.

The position was playable, even if theory said it was +2 for White. I'm going to show a position someone played against me, and in fact, my game was published in a ChessBase article , i won with black pieces

In another classical game there was also this 2100 FIDE-rated kid who had a personal coach a FIDE Master and both of them prepared a game at home against me, since they knew I was going to play the Latvian.

He came out of the opening with a +2 advantage, but that was only because I didn’t know that particular line back then.

Still, I ended up winning the game anyway (I was rated 2000 at the time, so he had 100 points on me), because the Latvian is a very dynamic opening and Black always has chances to create chaos.

I actually liked that weird Leonhardt line I had no clue what to play there, haha.

Of course, back then I also didn’t know about the Lichess database to study the line more deeply.

Those aren't such good openings

basically, jcidus entire argument is, he once caught a 2100 FIDE by surprise because he didnt prep enough agaisnt him. lmao.
+2 is GARBAGE. There is no sugar coating that. IT is a dead loss agaisnt sufficiently strong white opposition.

My argument is that the refuted line is still playable even in the worst case scenario, if your opponent goes into that variation you don’t want to face, there’s still hope of winning the game.

Back then, I studied it up to move 25 to understand the plans.

My opponent only knew it up to move 17 or 18, and assumed that since the engine said it was already winning, there was no need to look any further.

In vote Chess, the Killer Derivatives (a team I sometimes play on) beat the Unsound Openers in 44 moves. We had White in an Elephant Gambit. I'll give the first 10 moves: 1 e4, e5; 2 Nf3, d5; 3 exd5, Nf6; 4 Nxe5, Qxd5; 5 Qe2, Bc5; 6 Ng6+ (usually 6 Nc3 is played), Be6; 7 Nc3, Qf5; 8 Nxh8, Nc6; 9 d3, Nd4; 10 Qd2, O-O-O. My feeling is that it is not that bad for under 1800 USCF rated players in either blitz or very short time control games, ., G/45 5d (or less, of course). To play it in correspondence is downright suicidal and although my team did not always play the best moves (some moves were less than 3rd rate according to post game analysis), we still won. As Black I don't play double King pawn defences, yet I actually have a new found respect for this gambit despite the trash talk you often hear about it.

@Mazetoskylo Seeking to tap into your up to date knowledge here. I last looked into the Elephant Gambit when the Jensen, Purser & Pape massively enthusiastic chapbook on it was brand-new. And I couldn't see a line after 1. e4, e5; 2. Nf3, d5; 3. exd5, e4; 4. Qe2 which I felt really worked beyond having chances to fish in troubled waters. I experimented with 4..., Nf6 and 4..., Qe7 but had little joy with either. What do you feel is Black's playable fourth move?