Extending the same logic, a batsman with better average is definitely better than a batsman with lower average regardless of their SRs (in ODI).
So in ODIs, Dravid > Gilly.
This logic is sound in test matches, but not in ODIs.
Because in ODIs, ER is slightly more important than SR, while in test matches they are equally important.
That's why, average alone, can't tell you everything in ODIs, while it can reveal much in case of test matches.
This is true for both batsmen and bowlers.
Wrong. There's no connection between batsman's SR and their average unlike a bowler. In batsmen's case their Avg * SR would be a better measure. Don't bring in wrong concepts. ER and SR are both important. If you think that ER is very important look at the following.
Kapil Dev (01-10-1978 to 14-10-1994) - Avg - 27.45, ER - 3.71, SR - 44.2
Rest of the world (01-10-1978 to 14-10-1994) - Avg - 31.12, ER - 4.14, SR - 45.0
Kapil vs ROW: Avg: 1.13, ER - 1.12, SR - 1.02
Chaminda Vaas (15-02-1994 to 27-08-2008): Avg: 27.53, ER: 4.18, SR 39.4
ROW (15-02-1994 to 27-08-2008): Avg: 32.55, ER - 4.66, SR - 41.8
Vaas vs ROW:
Avg - 1.18, ER - 1.12, SR - 1.06
If the ER is the most important thing then Vaas > Kapil easily. Even if it was not ER, even Vaas' figures are better than Kapil's
That means Kapil was not the Asia's most potent bowler (High SR), he's not even the most economical, because he's on par with Vaas, but Vaas has a better SR and and average. As and ODI bowler Kapil is not even in the top 5 seamers of Asia (which i believe Wasim, Waqar, Shoaib, Imran and Vaas, perhaps Aaqib Javed as well)