Testing "my" "new" gambit idea vs. French defense

Recently I have been playing 1.e4 e6 2.d3 [pretending to play the KIA] then after 2...d5 3.Nf3 dx4 4.Ne5N
This move is not in the Lichess masters db (though 3.Nf3 is with a few games) but > 900 times in the Lichess players db (I've probably played < 30 games with it so far).
So far I have had good successes with it, also against 2200+ blitz rated opponents.
Here I had a pretty nice tactical game with this gambit :

Looks interesting... I spent quite a while about a year ago looking for gambit lines against the French (mainly because my results in the mainlines were pretty poor). I've never seen this before, though. What exactly would you say the compensation is for the pawn? Is it mainly just the opening of the d- and f-files?

In case you're interested, the gambit I play against the French at the moment is the wing gambit, which also can get quite tactical... it goes something like:
1. e4 e6 2. Nf3 d5 3. e5 c5 4. b4!? cxb4 5. a3 axb3 6. d4. The plan is to follow up with c3, creating a pawn chain that is pretty difficult for black to break down. I find it gives decent positional compensation, and also has a few traps that can catch unwary players out (especially in fast time controls). An example game (not the best example and it is a transposition from a Sicilian, but it shows some of the ideas): lichess.org/oyH7Uez72e7E My results so far against the French have been excellent... you can also play it against the Sicilian, but certain lines I've found aren't great for white.

Against the first gambit - instead of bd6 - what if black play Bc5? On first glance this threatens B.f2, if white retakes on e4 B.f2 too.
Waht would you like to play against 4... Bc5?
5. N.f7 could be met by Qf6

The wing gambit against the french - i don't llike that black could refuse the gambit and play d4. Should be ok for black with no great joy for white.

I'm not an FM, so my opponents don't play the best moves... but I did actually get a game where black refused the gambit. IMHO it's not bad for white... I ended up getting in a winning position (and this was in a decent time control OTB against a player who is 1900 OTB - so not a complete amateur) although I blundered later sadly. Unfortunately, I can't seem to find the game right now though...

Edit: I do actually have it: It's in a study here: lichess.org/study/ZKoM018t (round 5 game, I am white).

I don't would like to say that this is bad. But when i play gambit, i would like to get some fun. i don't like the sequence B:c5 and Qc7 in the opening for black, but i don't see something really better. Taking on c3 is a clear mistake in my opinion(ok, maybe inaccurate). After that white gets some fun in playing this game.
I woundered a little bit if black could play a gambit herself- b6 instead of B:c5. I think white is obliged to accept this and take the pawn on b6 and a7. Not sure about it... but with white playing a gambit i don't like black to turn the tables and get the iniative herself. I generraly don't like to take pawns when my opponent has enough compensation. I feel very unease then, more then i should do. I am not that bad in defending difficult positions, i simply don't like it.

@royalblue04
The move Bc5 looks very interesting, thanks.
It seems to be Lichess Stockfish their 2nd choice (I didn't let it run for long, and my computer is not very fast).
This gambit I play in blitz, and one advantage is that it takes the French player (I've played French myself as black) into unknown territory.
What I like in general is dynamic play, with piece activity over pawn grabbing.
When I am a pawn up myself I sometimes feel a slight pressure to win the game by keeping the pawn and shuffle and win slowly. In blitz that can cost quite some more time (And I never use pre move with this account, so I might get flagged).
I've had an opponent play f7f6?? as reply on Ne5.
I'll do some more experimenting and I hope that some opponents will try Bc5.

@davidharoldchess
Regarding wing gambit, I've played the a3 gambit vs. Sicilian in otb chess (with reasonable but not great results), and I must say I am not a huge fan of wing gambits though I still try them for fun in blitz on-line from time to time. When white doesn't get the desired piece activity, then one might have to suffer vs. the q side pawn majority of black.

@achja I'll definitely give your new gambit a go... I'm always on the lookout for interesting new ideas. As for what you said about the wing gambit, I agree that against the Sicilian it's usually not very good. I think this is because in the Sicilian wing gambit, black gets to choose what setup they will adopt. However, in the French black has already committed to a e6-d5 pawn structure. This means that you're pretty much guaranteed in the French wing gambit (assuming they don't play d4, which in my experience less than 5% of players do) to get a structure where white pawns are on c3-d4-e5 and black pawns are on d5-e6. In the French, black's usual plan is to play c5 to disrupt white's structure - but because black does not now have a c-pawn this plan is made impossible. To me, even just this fact makes the gambit appealing. In addition to this, white's pieces get active posts (Bd3, Nf3 and in many lines Ba3 Bxa3 Nxa3 followed by Nb5 is uncomfortable for black, with the opportunity to nestle in on d6). Castling kingside is usually dangerous for black due to the possibility of the Greek gift sacrifice.

@royalblue04 5... Bc5 is the move pretty much all masters play in the database (including a game by Viktor Kortschnoj from the 1970s). Qa5 is usually played instead of Qc7, but I failed to punish the move in the game anyway and it transposed. b6 looks interesting... only one game has been played so far with in the Lichess database. On the other hand, I wonder whether black has quite enough compensation for the two gambited pawns. For me all of this is moot though... I can count on one hand the number of games where black has played d4 instead of cxb4 or some other alternative.

5. f4 looks very bad after exf3 - you get worse version of blackmar-diemer after 6. Nxf3 and you have passive pawn on d3. If you are looking for something similar I would recommend the Krol gambit:
1. e4 e6 2. f4 d5 3. Nf3 dxe4 4. Ng5

@IAmMateCheckMate
5.f4 was indeed very bad in the game above.
I've actually played it several times, also later in the game where after 0-0 black had Qd4+ winning a pawn, and none of my opponents took advantage of it.
(However, I was very pleased with the double pin tactic + queen sac in that game)
Thanks for the Krol gambit suggestion.

@achja
You are just too good, congratulations for the play,
achja is a very dinnamic player his games are very nice to watch, sometimes I think what is the difference between a good player and an average player and It's usually badly answered since there is also the question of what is the difference of a bad player and an average player and in this case It's well answered, mistakes and blunders but the first question is very difficult to grasp and I think
It's related with the value of activity and the ability a player has to create opportunities in the game and this is the case of achja, a very nice player to watch and look at his games because I think the biggest mistake of an average player It's not errors or mistakes ironically but lack of ideas, lack of play, emptiness.
This is also why it is a big waste of time to play weak players if you have the opportunity to play stronger opposition as they teach you chess they teach you ideas, how to move the pieces, because It is very hard to learn chess in your own terms.
Trying to pick one idea at a time and experimenting with it like queenside castling in the french its cool practise to improve as a player and have more tools at your disposal to
surprise your opponent.