This is a good question, but I think it could use a slight reformulation: It's not so much a question of "how can we get our bishop to either d3 or c4", but rather "under which circumstances would the bishop become powerful on those diagonals".
If you're looking for positional ideas to further advance your repertoire in this line, it's often very useful to keep an eye on possible transpositions and openings that lead to analogous structures. For example, in the line you show, black has a very solid Slav structure, arguably even an improved one, since white doesn't have the Qb3 option anymore, so the Bg4 maneuver is perfectly sound. On the other hand, white's position has the same ingredients as most Larsen (1.b3) variants do.
As far as I remember (it's up to you to do the in depth search here), Larsen tended for the e3-f4 (instead of c4) pawn structure whenever he wanted to plant his bishops on b2 and e2. But he did have this Reti type of structure every now and then, see e.g. his game vs Ulf Andersson. So make sure you find out about the positional ideas Larsen used to employ in this system. But for a more extensive search, look into a database and search for all the games where this position has occurred and get inspired by how other players have gone about it. For example I use the TWIC database, I have a collection with many of the older series, and after a quick search I find about 200 games with this exact position.
What Bad_Bishop has suggested to you in his answer is actually a very practical idea, with a relatively clear plan ahead: you're threatening to win the bishop pair, and the weaknesses you may induce in doing so are less important than they seem specially because your weakening of the light squares is coupled with taking away the opponents light square bishop! Although it must be noted that in this specific line, after 9.Nh4, I'd be a bit worried about Be4, as f3 doesn't work due to Nh5. (Make sure you check this variation).
A couple of positional suggestions you might want to consider: (by no means exhaustive, in fact these will be very crude, just to give you some rough ideas)
A conservative approach: stick to the solid pawn structure and double the b2-g7 battery by bringing the queen to a1 or b2 and you coordinate your play around the control of the long diagonal. e.g. a sequence along the lines of: 7.d3 8.Nbd2 9.Re1 10.Qc2 11.Bc3 12.Qb2 etc should get you there... If you're not familiar with such battery ideas on the long diagonal, a good start would be to study this marvelous game by Peter Svidler playing as black: (notice that your pawn structure in this Reti line is completely transposable to that of Svidler's in this game):
[FEN ""]
[Event "Dortmund Sparkassen group-A"]
[Site "Dortmund GER"]
[Date "2004.07.26"]
[Round "5"]
[White "Arkadij Naiditsch"]
[Black "Peter Svidler"]
[Result "0-1"]
[ECO "B42"]
[WhiteElo "2574"]
[BlackElo "2727"]
[PlyCount "78"]
[EventDate "2004.07.22"]
1. e4 c5 2. Nf3 e6 3. d4 cxd4 4. Nxd4 a6 5. Bd3 Bc5 6. Nb3 Be7 7. O-O d6 8. c4 Nf6 9. Nc3 Nbd7 10. f4 Qc7 11. Qe2 b6 12. Bd2 Bb7 13. Rae1 O-O 14. Kh1 g6 15. Nd4 Rfe8 16. Nf3 Rad8 17. Ng5 Bf8 18. Qf2 Bg7 19. b4 Nh5 20. Rc1 Rc8 21. a3 Qd8 22. Nf3 Rc7 23. Rfe1 Nhf6 24. Qh4 Qa8 25. Qh3 Rec8 26. e5 dxe5 27. Nxe5 Nxe5 28. fxe5 Nd7 29. Bf4 Rxc4 30. Bxc4 Rxc4 31. Qg3 h5 32. Ne2 Nxe5 33. Rxc4 Nxc4 34. h4 Qd8 35. Bg5 Qd7 36. Nf4 e5 37. Nd3 Qd5 38. Nf2 Nd6 39. Rd1 Ne4 0-1
Go for the king asap: 7.Nc3 (idea being to force the retake on d5 to be by the e pawn and securing the f5 square for your f3 knight - with a clear attack on the kingside) 8.Nd4 ... Here's an example I could find:
[FEN ""]
[Event "70th ch-GER"]
[Site "Bremen GER"]
[Date "1998.11.12"]
[Round "7"]
[White "Hickl, J."]
[Black "Jussupow, A."]
[Result "1-0"]
[ECO "A12"]
[WhiteElo "2540"]
[BlackElo "2640"]
[PlyCount "59"]
[EventDate "1998.??.??"]
[SourceDate "2011.01.01"]
1. c4 c6 2. Nf3 d5 3. e3 Nf6 4. b3 Bg4 5. Bb2 e6 6. Be2 Nbd7 7. Nc3 Bd6 8. Nd4 Bxe2 9. Qxe2 O-O 10. cxd5 exd5 11. Nf5 Bc5 12. Na4 Bb4 13. Nxg7 b5 14. Nf5 bxa4 15. Qh5 Kh8 16. Qg4 Bxd2+ 17. Kxd2 Rg8 18. Qxa4 Rxg2 19. Ke2 Qg8 20. Ng3 Qg6 21. Rac1 Re8 22. Qf4 c5 23. Kf1 Rxg3 24. Bxf6+ Nxf6 25. hxg3 Ne4 26. Rh6 Qg7 27. Rd1 Re5 28. Kg2 Ng5 29. Rc6 h6 30. Rxh6+ 1-0
Transpose to a Queen's gambit decline type of position to secure the d3 square for your bishop: d4, Nbd2, h3, Bd3 ... Here's an example:
[FEN ""]
[Event "Golombek Mem"]
[Site "Paignton ENG"]
[Date "2000.09.05"]
[Round "3"]
[White "Bischoff, K."]
[Black "Hillarp Persson, T."]
[Result "1/2-1/2"]
[ECO "A12"]
[WhiteElo "2556"]
[BlackElo "2549"]
[PlyCount "37"]
[EventDate "2000.09.03"]
[SourceDate "2011.01.01"]
1. Nf3 Nf6 2. c4 c6 3. b3 d5 4. Bb2 Bg4 5. e3 e6 6. Be2 Nbd7 7. O-O Bd6 8.d4 O-O 9. Nbd2 Qb8 10. h3 Bh5 11. Bd3 Re8 12. Qc2 a5 13. Rac1 a4 14. Bc3 axb3 15. axb3 Bg6 16. Bxg6 hxg6 17. Ra1 Rxa1 18. Rxa1 e5 19. dxe5 1/2-1/2
Similar to Bad_Bishop's suggestion: start off with h3, make your opponent commit to a decision with the g4 bishop, but instead of an immediate rather committal g4, play the slow game: castles, d3, Nbd2, ... or possibly Nc3 specially if you plan to capture on d5 subsequently: two simple ideas to bear in mind: if black recaptures on d5 with the c-pawn, you want to take control of the c-file asap, if black recaptures with the e-pawn it's often possible to force the exchange of light square bishops in order to secure f5 for your knight. (I admit this is all still very crudely put).
Finally, note that within such Slav structures for black, you'll never see your bishop on c4, it's just incompatible with black's control of the center. Alternatively, if you want your bishop on d3, your best bet is to transpose (as early as possible) into Queen's pawn type of positions. To play openings such as the Larsen, English or the Reti, you really must like slow-positional plays, and in my opinion these are by no means easier to learn/play than open positions, in fact they can get extremely complicated rather quickly. Intuitively, this is because strategic ideas tend to be very abstract and hard to grasp, whereas open positions are more tactical which makes them inherently more concrete. To give you an example of a highly strategic game, played recently, quickly leading to a highly non-trivial middle-game (see the game below). For your purposes I highly recommend studying GM Michael Adams' games in the English opening.
[FEN ""]
[Event "GRENKE Chess Classic 2015"]
[Site "chess24.com"]
[Date "2015.02.07"]
[Round "5"]
[White "Adams, Michael"]
[Black "Bacrot, Etienne"]
[Result "1/2-1/2"]
[WhiteElo "2738"]
[BlackElo "2711"]
[PlyCount "129"]
[EventDate "2015.??.??"]
1. c4 Nf6 2. g3 g6 3. Bg2 c5 4. Nc3 Bg7 5. Nf3 O-O 6. d4 cxd4 7. Nxd4 Nc6 8. O-O Qa5 9. Nb3 Qh5 10. Nd5 d6 11. Nf4 Qe5 12. e4 Bg4 13. f3 Bd7 14. Rf2 b5 15. Nd3 Qh5 16. Nf4 Qe5 17. c5 dxc5 18. Nd3 Qc7 19. Nbxc5 Rfd8 20. Qe1 Be8 21. Bf4 e5 22. Be3 a5 23. Rc1 Qe7 24. Rd2 Nd4 25. Nb3 Nd7 26. Nxd4 exd4 27. Bf4 Rdc8 28. Bh3 Rc4 29. Rdd1 a4 30. Bd2 Ne5 31. Nxe5 Qxe5 32. Bf1 Rxc1 33. Bxc1 Qe6 34. a3 Rc8 35. Bd3 Qb3 36. Qe2 Bd7 37. Bf4 Be6 38. h4 Bc4 39. e5 Bxd3 40. Rxd3 Qd5 41. Kg2 h5 42. Qd2 Rc4 43. Qe2 b4 44. axb4 Rxb4 45. Kf2 Qb5 46. Rd2 Qd5 47. Qe4 Qa2 48. Qc2 Kh7 49. Kg2 Qa1 50. Rf2 a3 51. bxa3 Rb1 52. Qd3 Rd1 53. Bd2 Rg1+ 54. Kh3 Rh1+ 55. Rh2 Qg1 56. Qe2 Bxe5 57. f4 Bd6 58. a4 d3 59. Qg2 Rxh2+ 60. Qxh2 Qd1 61. Qg2 Ba3 62. a5 Bc1 63. Qd5 Qf1+ 64. Kh2 Qe2+ 65. Kh3 1/2-1/2
