I confess I've never been clear about when dxc5 might be good in Advance French style positions -- I have the impression this can be a tough one even for very strong players! But here after 7 ...Bc5 8 0-0 f6 9 Qe2 fe Black seems to do OK, though there are very few games. (The engine much prefers this for Black to 7 0-0 Rc8 8 dc, which might indicate my III(a) line above is inferior?)

Incidentally ChessPublishing's own Michael Roiz has had this position (7 dc) as Black, the game continuing 7 ...Bc5 8 Nbd2 f6.

The O’Kelly Sicilian (1 e4 c5 2 Nf3 a6) is sometimes seen as inferior, even dubious, on account of 3 c3!. But if, after 3 …e6 4 d4 d5, Black is happy with 5 ed ed (after which AFAIK best for both is Emms–Glek but with 12 …Bf5 or 12 …Re8), what, I ask, is really inferior about 5 e5 Bd7? Unless 6 g3 is a hugely serious try (6 …cd!? 7 cd Bb5; 6 …Nc6!?), it seems play is headed for a 5 …Bd7 Advance French where …a7-a6 may be rare, but is only very minimally inferior to more common lines, if that. For instance: 

I  6 a3 Nc6 (Andrew Martin gives 6 …Qb6 7 b4 cd 8 cd Bb5!) and:

(a)      7 Be2 Rc8 8 0-0. Here 8 …cd 9 cd Nge7 is solid, while Morozevich has successfully played 8 …h6!?
(b)      7 Bd3 Rc8 (also 7 …Qb6 7 Bc2 Nge7/Nh6 may be OK?) transposes to an AF with 5 …Bd7 6 Bd3 Rc8 7 a3!? a6, e.g. 8 0-0 cd 9 cd Qb6 10 Bc2 g5!
(c)      7 b4 c4!? (or 7 …cd 8 cd Nh6/Nge7)

II  6 Be2 Nc6 (with Bc2 now impossible 6 …cd 7 cd Bb5 may be OK too?) 7 0-0 Rc8 8 Nbd2 (8 a3 is I(a) above, while 8 Na3 [or Be3] cd 9 cd Nge7 is fine for Black) cd 9 cd Nge7 (or 9 …Qb6 10 Nb3 Nb4!?)

III  6 Bd3 Nc6 (strangely perhaps, this move order appears to arise only via the O’Kelly) 7 0-0 (7 a3 is I(b) above) and:

(a)      7 …Rc8 (Glek). Now 8 a3 cd is I(b) above, but 8 dc!? Bc5 9 b4 may be messily +/=?
(b)      7 …cd 8 cd Qb6 9 Nc3 (9 Be2 Nge7) Nd4 10 Nd4 Qd4 is the good old Milner-Barry Gambit!

It seems to me that if White can get any distinct advantage here at all, the way to do it is not so obvious. Does anyone know more?