3e was not destroyed by magic item 'shopping'.
The high end game balance was affected by the multiple access to layering bonuses. Magic weapons had a bonus, buffs to attributes had a bonus, polymorph or wild shape changes had a bonus, divine bonus, luck bonus, feat bonus, epic bonus, and more.
Further, it was not just 'shopping' for magical items but it was item creation. You could mostly sit at home and make money manufacturing magical items as the normal return on materials was 2 to 1. Start with 50 gp and make two level 1 potions of healing. Sell one potion for 50gp and pocket the other as profit. Rinse and repeat for the infinite supply of healing potions.
Down time means very little to players as months can often go by in an hour of descriptive time (though many players seem to be concerned that every hour is accounted for to get maximum advantage). It does not take long to prove an infinite economy of resources for most mages and clerics with little need to adventure.
Stores are not needed in 3e/PF as most players that do not stumble across the item they want will make it themselves with usually a profit if they make more than one. Need a belt of +4 Strength then use some gold to make it or borrow the money and make two with the sale of one belt to cover the debt.
The first lesson of magical creation should be that the reverse ratio or worse is the proper ratio. If it costs 100 gp to make a potion of healing by yourself and the normal market price for the item is 50gp (do to factors maybe of larger producers of the item and scale) then players won't produce the infinite economy and will tend to buy what they need or only produce what they need.
---------------------
A person brought up that they felt that they should have an easier time fighting a giant with a +4 weapon then without or if they had leveled up.
I had mentioned that the level increase of a party tends to result in a no net change in difference between opponents and players as the GM tends to 'level' up the monsters defenses and hazards to match the level increase of the players. It is the level tread mill that results in no net increase but a large amount of mathematical totaling on both sides of the screen which just nets out to roughly zero in the best case or results in a wide variation in the player's abilities to handle a threat (I've seen differences of +10 to +20 in to hit and AC in the late teens of PF and 3.5 between players).
I proposed a removal of the level increase for a set of static levels of play that cover choices of complexity that players like to use. Players that like to have few choices can choose one level of complexity and players that like to have twenty or more choices on their turn can have their complexity level.
This leads back to the giant which the player will always find is a giant and hard to fight even with items.
What I think this then results in the DM moving the level mechanic to the monsters and just deciding how much tougher a monster is in relation to the players. You could have an 'easy' giant and a 'hard' giant. The giants could be similar but with a variation of defenses and hit points they would feel different. The GM could reward story play by swapping in an easier giant if the players have taken steps to prepare or have certain items.
