As for recent speakers, I vote for the Infinity R152's and R263's! Like their predecessors in the Primus lineup (the 150 and 360), these are the super-achievers in the Reference family. Also, I haven't heard the R253's but the Harman spinorama measurements look very good as well. One of these three models should get very close to the more expensive Revels.
For instance, the R152 achieves an Olive Preference rating (5.3) very near that of the Revel M16 (5.5). (Source:
https://pierreaubert.github.io/spinorama/Infinity R152/Infinity/index.html ), although in practical listening these two speakers provide somewhat divergent presentations. Interpreting the spinorama data helps explain why (and shows why a full set of measurements is more useful to predicting sound than just a single preference score). On one hand, there is the midbass richness of the M16 and, on the other hand, the slightly elevated top end of the R152. What this translates into during daily-driver listening is that the R152 behaves more like a near-field minimonitor, whereas the M16 sounds more like a floorstander squeezed into a bookshelf form-factor. The R152 feels a bit faster and more resolving of transient information, while the M16 sounds way bigger and effortless than its size might suggest. I suspect this is because the R152 lacks the bass-frequency masking of the M16, while the M16 is a lower-distortion configuration that stays linear (same frequency response) through a wider SPL range.
Getting a handle on just how far we can trust the Olive Score and Spinorama measurements as predictors of sound, and to what extent they fall short, would be aided by testing out the broader Reference line, as they all are: 1) Inexpensive to acquire, 2) Designed to score high on the Olive Factor, and 3) Generally lacking in serious design defects, being typically-Harman in their thoroughness of design. Comparing these to the more expensive Revel lineup and understanding their similarities and divergences would be instructive.
Also, scouting out and testing older Audiophile Classics such as the Spica TC-50, Snell K/II, and the Dunlavy SC-I would be interesting. Those speakers represent budget state-of-the-art circa the mid-1990's and got all the rave (subjective) reviews back then. It's 30 years later and we've come a long way since then thanks to the science of testing, measurement, materials science, and computer-aided optimization of design, to say nothing of lower foreign manufacturing costs.
More recently, in the mid-2000's, the PSB B25 and T55 speakers were really something special in the budget domain. And, of course, the fascinating Primus 360 (not the P362 or P363, which tipped up the tweeter output on-axis to achieve a flatter in-room response).
We could learn a lot by testing out and comparing these well-designed budget speakers and pointing out those where the stars do seem to align and produce a bulletproof listening experience (as I feel the Primus 360 and Revel M16 do).