This one of the more detailed examinations of draft pick values.
If you look at this chart, the drop-off in value from #3 to #5 is roughly the same as the drop-off in value from #5 to the end of the 1st round.
Unfortunately, the chances of a "generational" talent at #5 are not good, and substantially less than at #3.
I'll be the first to admit I'm not strong on statistics, but it seems to me from the article that the chance of drafting a top player is higher with two mid-round picks than with one 5th overall. For example in 2013, a #9 pick (BoHo) and a #23 pick (Burakovsky) are more valuable than one 5th overall (Elias Lindholm).
If you can convert the 5th overall into a combo of picks that gives a greater statistical probability of getting a top forward (e.g. 2 late-1sts and a 2nd, 2 mid-1sts), then I think JB has to go for it. Its not as though drafting in the mid- or late-1st round guarantees the player will be a bottom-six forward.
And when it comes to defenseman, the chances of getting a top-4 in later picks goes up relative to forwards, so if that is an organizational need, it makes sense to get more picks rather than high picks.
Now if you could get Montreal's 9th, and one (or both) of their seconds, for Dubois, that would yuuge.