Ah yes, the old intellectually dishonest argument that unless you're willing to draft the pure BPA every single time no matter the circumstance then you must draft strictly based on need, very nice.
Need of course factors in but it's a sliding scale that needs to be constantly re-calibrated based on the position in question, your roster, and the talent available. If a WR is far and away the BPA in round 1 or 2 and we haven't signed a Cobb-replacement, you take him, because that player could have a huge role in 2020 and moving forward if we decide to move on from Amari after two years.
If there is an equally rated player at a greater positional need, or a player that is rated just a notch lower, then the conversation changes obviously, but if a GM were to completely write off positions in certain rounds because they don't think they're as needy as others then that's a good way to building a shit roster long-term.
There are limits to it of course, nobody is going to draft a swing OT in the 1st with two Pro Bowlers in their prime or a backup QB to sit behind a 25-year old Pro Bowl QB. It's much more complex of a decision when you're talking about a 3rd WR who would likely play at least 50% of the snaps and who could potentially be your top WR in 2-3 years, or an edge rusher who would be a big part of your nickel/dime and rotate in during base downs as a rookie but who you project to be maybe your top edge rusher in 2 years.