[shorter version of my initial post for a quicker read]
First off, nothing should be set as an absolute in fantasy football as you should always adapt to how a draft is going, so let me preface the analysis by saying that.
To narrow down the analysis, I'm going to specifically base my argument on a person drafting 1st, choosing LT, and hence, having picks 20 and 21 in rounds 2 and 3. While I think the strategy extends out to more RBs (guys like Westbrook, ADP, etc.) and draft positions, it's easier to do the analysis simplified to this one set of circumstances. And more importantly, I think most of the people who argue against the merits behind the strategy of RB/WR/WR would still have contentions with it in this situation.
To start off, I'm going to list the guys who are almost assuredly going to be drafted prior to pick 20:
RB: LT, Westbrook, ADP, S Jax, Addai, Gore, MB III, LJ, Portis
WR: Randy Moss, TO
Which is 12 players, leaving 7 more to be drafted prior to pick 20. I'd expect at a bare minimum, a few more RBs/QBs to be drafted, and as long as 3 are, then there's no deviation from the strategy. I also see what is a pretty clearly defined tier of WRs behind Moss and TO, consisting of Wayne, Braylon, Colston, Andre, Fitz, and Steve Smith.
Which brings us to picks 20 and 21. With the 12 players listed above being drafted as well as 3+ additional RBs/QBs, at least 2 of these wideouts from this 6 WR tier will be available and should be taken. The one caveat I'll mention off the bat is Reggie Bush, who I view as a wildcard if available, and I could see justifying taking him along with a wideout here instead of 2 WRs. For the sake of this analysis, I'm going to discount him and leave that as a separate decision for owners to make.
1.) The first comparison to be done is a heads up scoring comparison between the available RBs and WRs. Based on my projections with all players playing 16 essentially healthy games, the 6 WRs project to around an average around 295 points with nobody under 284. I'll use 290 as a basis as the 5 outside of Steve Smith average closer to 290 and his value should come down a notch from the 300+ projection. Comparably, I don't have a single potentially available RB (Bush excluded) who I project higher than 270 points and the average of this grouping of RBs is around 255. Pretty clear-cut advantage in straight up points value here.
2.) The second comparison is the likelihood of the RBs available and the WRs available to perform within/exceed expectations versus falling short, be it due to injury, under-performing on the field, whatever. I don't have time to go back through all the drafts of recent years, but I'm nearly positive the last several seasons consistently favor the chances of a WR from the comparable tier performing within/exceeding expectations than a comparable 2nd tier RB. Here's a post I made in 2006 comparing 2nd tier RBs vs top tier WRs (slightly different analysis as the wideouts and RBs discussed are higher than those discussed for this strategy as the other post was for WR/WR in rounds 1 and 2 if not getting a top/middle draft position): viewtopic.php?f=1&t=230131&p=1635224#p1635224
3.) Lastly, let's compare the drop-off in expected point production for RBs and WRs available in later rounds. First, let's look at rounds 4 and 5 (picks 40 and 41). The best RBs who may potentially be available are the likes of Parker, Maroney, Graham, Ronnie Brown, T. Jones, Edge, etc. Guys like Maroney and Ronnie are difficult to project and have high upsides and downsides whereas the others can reasonably be projected between say, 220 and 260 points. I'll go with a conservative value of 225 points in assigning the value of an available RB here.
As for the best potentially available WRs, I'd include the likes of the Detroit wideouts, Santonio, Harrison, Cotchery, and Marshall. I have these guys projected out to between 220 and 265 points (with the exception of Marshall who I have a little higher over 16 games but has the suspension). The values average out to around 240 points for these guys.
So you're looking at a drop-off in about 30 points if waiting on a RB vs. about 50 points for a WR from rounds 2/3 to 4/5. As you go into the middle rounds of 6-9, there's naturally a starker drop-off at the RB position so you may choose to draft 2 RBs in rounds 4 and 5 if going WR/WR in 2/3 although it's not a necessity to do so.