YES

Problem:
 a__f(a(),X,X) -> a__f(X,a__b(),b())
 a__b() -> a()
 mark(f(X1,X2,X3)) -> a__f(X1,mark(X2),X3)
 mark(b()) -> a__b()
 mark(a()) -> a()
 a__f(X1,X2,X3) -> f(X1,X2,X3)
 a__b() -> b()

Proof:
 Matrix Interpretation Processor: dim=1
  
  interpretation:
   [f](x0, x1, x2) = 5x0 + x1 + 4x2,
   
   [a__f](x0, x1, x2) = 5x0 + x1 + 4x2,
   
   [b] = 2,
   
   [mark](x0) = x0 + 2,
   
   [a] = 2,
   
   [a__b] = 2
  orientation:
   a__f(a(),X,X) = 5X + 10 >= 5X + 10 = a__f(X,a__b(),b())
   
   a__b() = 2 >= 2 = a()
   
   mark(f(X1,X2,X3)) = 5X1 + X2 + 4X3 + 2 >= 5X1 + X2 + 4X3 + 2 = a__f(X1,mark(X2),X3)
   
   mark(b()) = 4 >= 2 = a__b()
   
   mark(a()) = 4 >= 2 = a()
   
   a__f(X1,X2,X3) = 5X1 + X2 + 4X3 >= 5X1 + X2 + 4X3 = f(X1,X2,X3)
   
   a__b() = 2 >= 2 = b()
  problem:
   a__f(a(),X,X) -> a__f(X,a__b(),b())
   a__b() -> a()
   mark(f(X1,X2,X3)) -> a__f(X1,mark(X2),X3)
   a__f(X1,X2,X3) -> f(X1,X2,X3)
   a__b() -> b()
  Matrix Interpretation Processor: dim=3
   
   interpretation:
                      [1 1 1]     [1 0 0]     [1 1 1]  
    [f](x0, x1, x2) = [0 1 0]x0 + [1 0 0]x1 + [0 0 0]x2
                      [1 0 0]     [1 0 0]     [0 1 0]  ,
    
                         [1 1 1]     [1 0 0]     [1 1 1]     [0]
    [a__f](x0, x1, x2) = [1 1 0]x0 + [1 0 0]x1 + [0 1 0]x2 + [1]
                         [1 1 0]     [1 0 0]     [0 1 0]     [1],
    
          [0]
    [b] = [0]
          [1],
    
                 [1 0 0]     [0]
    [mark](x0) = [1 1 0]x0 + [1]
                 [0 1 1]     [1],
    
          [0]
    [a] = [1]
          [1],
    
             [1]
    [a__b] = [1]
             [1]
   orientation:
                    [2 1 1]    [2]    [1 1 1]    [2]                     
    a__f(a(),X,X) = [1 1 0]X + [2] >= [1 1 0]X + [2] = a__f(X,a__b(),b())
                    [1 1 0]    [2]    [1 1 0]    [2]                     
    
             [1]    [0]      
    a__b() = [1] >= [1] = a()
             [1]    [1]      
    
                        [1 1 1]     [1 0 0]     [1 1 1]     [0]    [1 1 1]     [1 0 0]     [1 1 1]     [0]                       
    mark(f(X1,X2,X3)) = [1 2 1]X1 + [2 0 0]X2 + [1 1 1]X3 + [1] >= [1 1 0]X1 + [1 0 0]X2 + [0 1 0]X3 + [1] = a__f(X1,mark(X2),X3)
                        [1 1 0]     [2 0 0]     [0 1 0]     [1]    [1 1 0]     [1 0 0]     [0 1 0]     [1]                       
    
                     [1 1 1]     [1 0 0]     [1 1 1]     [0]    [1 1 1]     [1 0 0]     [1 1 1]                
    a__f(X1,X2,X3) = [1 1 0]X1 + [1 0 0]X2 + [0 1 0]X3 + [1] >= [0 1 0]X1 + [1 0 0]X2 + [0 0 0]X3 = f(X1,X2,X3)
                     [1 1 0]     [1 0 0]     [0 1 0]     [1]    [1 0 0]     [1 0 0]     [0 1 0]                
    
             [1]    [0]      
    a__b() = [1] >= [0] = b()
             [1]    [1]      
   problem:
    a__f(a(),X,X) -> a__f(X,a__b(),b())
    mark(f(X1,X2,X3)) -> a__f(X1,mark(X2),X3)
    a__f(X1,X2,X3) -> f(X1,X2,X3)
   Matrix Interpretation Processor: dim=1
    
    interpretation:
     [f](x0, x1, x2) = x0 + 4x1 + x2 + 2,
     
     [a__f](x0, x1, x2) = 5x0 + 4x1 + x2 + 4,
     
     [b] = 7,
     
     [mark](x0) = 5x0 + 2,
     
     [a] = 3,
     
     [a__b] = 2
    orientation:
     a__f(a(),X,X) = 5X + 19 >= 5X + 19 = a__f(X,a__b(),b())
     
     mark(f(X1,X2,X3)) = 5X1 + 20X2 + 5X3 + 12 >= 5X1 + 20X2 + X3 + 12 = a__f(X1,mark(X2),X3)
     
     a__f(X1,X2,X3) = 5X1 + 4X2 + X3 + 4 >= X1 + 4X2 + X3 + 2 = f(X1,X2,X3)
    problem:
     a__f(a(),X,X) -> a__f(X,a__b(),b())
     mark(f(X1,X2,X3)) -> a__f(X1,mark(X2),X3)
    Matrix Interpretation Processor: dim=3
     
     interpretation:
                        [1 0 0]     [1 0 1]     [1 0 0]     [0]
      [f](x0, x1, x2) = [0 0 0]x0 + [1 1 1]x1 + [0 0 0]x2 + [1]
                        [0 1 1]     [0 0 0]     [0 0 1]     [1],
      
                           [1 1 1]     [1 1 0]     [1 0 1]     [0]
      [a__f](x0, x1, x2) = [1 1 1]x0 + [0 1 0]x1 + [1 0 1]x2 + [1]
                           [0 0 0]     [0 0 0]     [0 0 0]     [1],
      
            [0]
      [b] = [0]
            [1],
      
                   [1 1 1]     [1]
      [mark](x0) = [1 0 1]x0 + [1]
                   [0 0 0]     [1],
      
            [1]
      [a] = [0]
            [0],
      
               [0]
      [a__b] = [0]
               [0]
     orientation:
                      [2 1 1]    [1]    [1 1 1]    [1]                     
      a__f(a(),X,X) = [1 1 1]X + [2] >= [1 1 1]X + [2] = a__f(X,a__b(),b())
                      [0 0 0]    [1]    [0 0 0]    [1]                     
      
                          [1 1 1]     [2 1 2]     [1 0 1]     [3]    [1 1 1]     [2 1 2]     [1 0 1]     [2]                       
      mark(f(X1,X2,X3)) = [1 1 1]X1 + [1 0 1]X2 + [1 0 1]X3 + [2] >= [1 1 1]X1 + [1 0 1]X2 + [1 0 1]X3 + [2] = a__f(X1,mark(X2),X3)
                          [0 0 0]     [0 0 0]     [0 0 0]     [1]    [0 0 0]     [0 0 0]     [0 0 0]     [1]                       
     problem:
      a__f(a(),X,X) -> a__f(X,a__b(),b())
     Matrix Interpretation Processor: dim=3
      
      interpretation:
                            [1 0 1]     [1 0 1]     [1 0 0]     [0]
       [a__f](x0, x1, x2) = [1 1 0]x0 + [0 0 0]x1 + [1 1 0]x2 + [1]
                            [0 0 0]     [0 0 0]     [0 0 0]     [0],
       
             [0]
       [b] = [1]
             [0],
       
             [0]
       [a] = [1]
             [1],
       
                [0]
       [a__b] = [0]
                [0]
      orientation:
                       [2 0 1]    [1]    [1 0 1]    [0]                     
       a__f(a(),X,X) = [1 1 0]X + [2] >= [1 1 0]X + [2] = a__f(X,a__b(),b())
                       [0 0 0]    [0]    [0 0 0]    [0]                     
      problem:
       
      Qed