I think 1. d4 d5 2. c4 e6 3. Nc3 Nf6 4. cxd5 exd5 5. Bg5 c6 6. e3 Bd6 qualifies as an unusual line. I can't say it's new, with a couple games from the 1940s and many in the 1950s. It has been played many times but it has only a fraction of the number of games compared to positioning the bishop on e7. The immediate comparison is evident. ...Bd6 is an active square but ...Be7 breaks the pin.
6...Bd6 has received recent attention. Davies, Chess Publishing November 2025, devotes an entire issue to the move. (Illingworth also covered it, May 2017). It has received a good scrub from Arnaudov, Modern-Chess August 2025, and Rodi, Forward Chess 2024. Semkov looked at it from the White side, Forward Chess 2025.
6...Bd6 leads to a different kind of game compared to ...Be7 lines. Despite the attention it has received it remains very open-ended with lots of lines yet to be discovered.
1. d4 d5 2. c4 e6 3. Nc3 Nf6 4. cd5 ed5 5. Bg5 c6 6. e3 Bd6 7. Bd3 O-O 8. Nge2 (Semkov advocates 8.Nf3!?) Re8 9. Qc2 h6 10. Bh4. Here Arnaudov advocates 10...a5. Davies examines 10...a5 in all but one game. Rodi considered 10...a5, 10...Nbd7, 10...Be6, and 10...b6.
After 10... Nbd7 11. h3, Rodi says that black has played many moves, including 11...Be7 and 11...Qa5, and he advances 11...Nf8 as a logical continuation.
I decided to look at 11... b5 12. O-O a5. Stockfish suggested the direct 13. e4. (Upon further exploration, I notice that humans have done well with 13.Rc1, 13.Rad1, and 13.Rae1, winning one game with each move. So that deserves a look.). Continuing on, Stockfish is quite optimistic about white's chances until it comes to one of those recalibration moments. 13...g5 14. Bg3 Bg3 15.Ng3 b4 16. Na4 de4 17. Be4 Ne4 18. Ne4 Nf8 19. Qc6 Ra6 20. Qc2 g4 where black has compensation for his pawn.