I would not say additional moves are illegal but the game should automatically be considered drawn at that point. FIDE expects that a game in that situation is being observed by an arbiter and the arbiter will declare it drawn at that point.
75 move rule and checkmate
Fivefold repetition is harder.
Say, an endgame like KQ vs. KR. For a tablebase, a forced checkmate exists in under 50 moves. For a human, it may be hard to find the best moves.
Same material, no pawns to move and no captures as yet. The queen is trying to find the forced win or spot and exploit rook´s blunders. The rook is trying not to blunder, exploit queen´s blunders to stave off forced checkmate, and not run out of time.
Checking if the rook is yet safe at 50 moves is relatively easy. Pawn moves are noted on scoresheet and as per laws:
"C.9 When a piece makes a capture, an x may be inserted"
"C.13 Abbreviations
...
x = captures
...
The last four are optional."
Noting captures, while optional, is allowed, and useful - to quickly scan scoresheet for captures or pawn moves, and check the number of moves without pawn moves or captures.
But repetition of positions?
Positions are not expressly written on scoresheet. Even if they were, pairwise checking 100 positions for duplicates takes time.
There may be deja vu. But if a position has occurred through different moves, like move 5, and 45, would the player be sure whether it´s just 2nd time, so no claim, or it also happened on move 20? False claim is expressly penalized. Checking the scoresheet takes time.
If the position has recurred 3 times, then the right to claim exists until the player on the move touches a piece with intent to move. But 5-fold repetition is an automatic draw.
So, if the same position existed, via different moves each time, at moves 5, 15, 25, 35 and 45 (where the rook, for reasons offered above, does not notice to be sure of claim either on move 25, 35 or 45), and the rook was finally captured on move 49, soon followed by a forced checkmate KQ vs. K... could rereading the scoresheet after the checkmate, or rook finally running out of time after move 45, or resigning after move 45, show that a 5-fold repetition did happen, and the game was drawn?
If both players fail to notice the automatic draw, and sign the scoresheets, can the arbiter on noticing the fivefold repetition impose the draw?

The 5 repetition rule states consecutive alternate moves, so, repeating the position for a 5th time won't trigger the automatic draw say if the 3rd was on move 30, 4th on move 34, and 5th on move 40.
I believe it is to stop people from shuffling back and forth while not claiming triple-repitition or while it can't be caimed (no notation for example), or each player is trying to flag the other.
So, while the rules are related, they don't trigger in the exact same manner.
The rule is:
"9.6
If one or both of the following occur(s) then the game is drawn:
b. any consecutive series of75 moves have been completed by each player without the movement of any pawn and without any capture. If the last move resulted in checkmate, that shall take precedence."
So checking the moves... perhaps it´s better to refer to "ply"?
Numbering the last ply that was pawn move or capture as ply 0:
the first ply on which a right to claim exists is ply 100, and then by writing down intended move and calling arbiter
the first ply on which the draw can be claimed directly from pre-move position is ply 101
the last legal move is ply 150
if ply 150 is a checkmate, it stands even if it was not a capture or pawn move (nor is a checkmated player having the move)
all moves from ply 151 on are illegal, and if the game ends by checkmate later, it is retracted.
Correct?