This is not an answer. It is just the somewhat trivial observation that WLOG you can relax the requirement that there be exactly p edge subsets {Ei}i of exactly the same size, and instead just look for any number of edge subsets of of size O(the desired size). Maybe this helps think about the problem.
Fix any graph G=(V,E) and integer p≥1. Let s=⌈|E|/p⌉
Lemma. Suppose there are subgraphs {G′j=(V′j,E′j)}j such that {E′j}j partitions E into (any number of) parts of size O(s).
Let M=maxv∈V|{j:v∈V′j}|
be the maximum number of parts that any vertex is in.
Then there are p subgraphs {Gi=(Vi,Ei)}i such that {Ei}i partitions E into exactly p parts each of size at most
s=⌈|E|/p⌉, and
maxv∈V|{i:v∈Vi}|=O(M).
Proof. Starting with the sequence E′1,E′2,…,E′p′, replace each part E′j in the sequence by any ordered sequence of the edges contained in that part. Let e1,e2,…,em be the resulting sequence (a permutation of E such that each part E′j is some "interval" {ea,ea+1,…,eb} of edges in the sequence). Now partition this sequence into p contiguous subsequences such that each except the last has size s, and let Ei contain the edges in the ith contiguous subsequence. (So Ei={eis+1,eis+1,…,e(i+1)s} for i<p.)
By assumption each part E′j has size O(s), and by design each part Ej except the last part Ep has size s, so (because of the way {Ei}i is defined) the edges in any given part E′j are split across O(1) parts in {Ei}i. This, and the assumption that each vertex occurs in at most M of the parts in {E′j}j, imply that each vertex occurs in at most O(M) of the parts in {Ei}i. QED