Ron: After 3. ... c6 my Stockfish 7 preferred c4. Sandu played Nc3
After 4. ... d6 my Stockfish preferred a4. Sandu played Be2
5. 0-0 by Sandu is a commonsensical developing move.
After 5. ... 0-0 my Stockfish preferred a4, as played by Sandu.
After 5. ... Qc7 my Stockfish prefered, after some time h3, as played by Sandu
Moves in parentheses is move by Sandu, Stockfish moves in non-parentheses
9 (Bf4) Bf4
10. (Be3) Bg5
11. (dxe) Qd2
12. (Ne2) Nb1
13. (Nc4) Nc4
14. (Nb6) Nb6
15. (Qxe2) Qe2
16. (Nc4) Nc4
17. (axb5) axb5
18. (Nd2) Nd2
19. (Nd5) Nd5
20 (exd5) exd5
21. (Nb3) Nb3
22. (Bc5) Bc5
23. (Qe3) gxf3
20 Qe3 was a blunder by Sandu, the evaluations change from advantage White to advantage Black
24. (Rd1) Re1
25. (d6) d6
26. (Ba7) Qe3
27. (Na7) Na7
Here, Stefanova blunders with 27. ... Qb5. White eval, which was about zero, now jumps to over 7
28. (Qb3+) Qb3
29. (c4) c4
Notice that Black is forced to lose material.
I'm not going to go further. In my judgment, Sandu did not use computer assistance. There was variance in the opening moves that Sandu played. Sandu's moves that matched a computer were commonsensical moves such as castling, recapturing, and moving an attacked piece to a good square.
One move by Sandu had flipped the advantage from her to her opponent.
The position after 27. .. Qb5 would make a good puzzle of the week which at least some members of chess games.com would solve.
So what do you have to say, <BloatedBelly>?