5e's balance system is very, very rough.
When it works, it gets you in the right ballpark, and it attempts to give you the tools to understand how much harder a larger encounter is, so you aren't suprised.
But 5e's balance system is not nearly tight enough to distinguish between a fight where everyone has to stretch to win, and one where the entire party is wiped. Or even between a routine fight and a stretch fight.
So regardless of how much gear you give players, you must experiment with how combat capable your party is. If they win an EL+3 encounter without a single character being KO'd or significant daily resources being burned, you'll know that EL+3 isn't super-hard encounter for them. On the other hand, if they burn almost all of their daily resources and most of the party is KO'd during a EL+3 encounter, you'll know EL+3 is a super-hard encounter for them.
Make sure you aren't getting the math wrong (5e encounter math is a bit fiddly), throw EL+0 encounters at them, and then throw a +1 or +2 and see how they do. If those are still easy, throw +2 or +3s at them as "hard" encounters. If those are still easy, throw +3s or +4s as the hard encounter (mixed in with +0s still).
A bunch of good magic items can make a difference, but the character builds and skill of the players can make a larger one; as can the DM's tactical choices when running team monster. A party that leaves squishies exposed, uses weak combat builds, wastes daily resources, and a DM that viciously expoits team-PC flaws will have a much larger balance impact than a full set of +3 gear. So simply stating "X magic items means CR goes up by Y" doesn't work well.
Now, in a few cases, I've wanted to have a really close fight. For example, a duel between a Paladin and an Arch-foe. I've actually gone and simulated the fight to determine how close it will be, as I didn't want a blowout in either direction. Even a simple technique, like tracking each sides at-will DPR against a simple foe model, can be used to ensure a fight won't be a complete blowout.