Doing this more comprehensively now
Charizard: Flamethrower's ending animation becomes land-cancelable. One of Charizard's biggest problems is and always has been engaging from an air-to-ground transition. Even something as slow and predictable as Flamethrower has multiple hits, and letting it land cancel would even be a unique attribute. There are only three other short-range projectiles of a similar nature in the game, and they already have unique aspects: Squirtle's Down-B give him a little hop and trips, Ice Climbers can desynch blizzards for a Fence of Pain, and Bowser can Flame Cancel the start-up of his. Right now, Charizard's Neutral B is trying to be not Bowser's by the mechanics of the flame itself -- Aimable, further reaching, etc. -- but without any unique function to the move, it just falls flat and feels like an inferior attack in spite of the differences already in place.
Yoshi: After careful consideration of changes to his moveset, I came to the conclusion that it is actually fine. However, a touch more aerial acceleration and stopping friction would do wonders for his recovery and on-stage harrasment with, particualarly, bair. His air speed is fine, but a little move weaving action would be beneficial.
Zelda: Again, let Zelda air dodge out of Side B.
Pikachu: Again, Smash 64's large, cancelable Fair.
Squirtle: I think it would be a safe move to let Neutral B deal damage at a distance. It already decays to 1% before much distance; having it deal damage with the non-flinch pushback would be perfectly adequate for move that sees such limited use anyway.
Ness: Fix dash attack. Above all other changes Ness could want, this is, upon thought, the one that stands. The hits don't link together, the final hit stops having anything resembling combo capability after 60%, and the move is, besides terribly telegraphed, completely unsafe on HIT, let alone shield, mostly do to the non-linking hits, but also due to its susceptibility to CC'ing. Simna had a tendency to call the move "PK Shove" in Melee, due to its scaling knockback on the initial hits that got to a point that it just pushed the opponent off the stage horizontally, going past the second and third hits entirely. If this aspect was played up more, it could help the move find use in Ness's gameplay. That is to say, let the horizontal hits scale in knockback. They'll link into the final hit at lower percents as they should, but the function changes quickly as knockback scales and it becomes a move that quickly puts the opponent in a recovery spot.