Interesting how many comments are about the R3's bass.
I realize integrating subs (and buying subs) can be a bit of a pain, that said there is no comparison between any of my bookshelf speakers with and with out subs.
Subs beat no subs every time bar none.
Now that I have my way to integrating subs down to a personal science, I have never once thought my speakers sound better without subs.
I listen to a wide variety of music, I do listen to quite a selection of tunes with bass such as all sorts of electronic stuff. I also listen medium loud to louder regularly.
R3's need subs, @least in my 12x23 foot bigger side of medium room. Not even close, even with small subs.
Even if I get a similar moving mic in room energy trend the difference is immediate and I have to recommend subs here.
Human hearing is just is not sensitive to bass and needs extra SPL there as well so subs allow for that to be adjusted in way that is difficult with just the monitors. I think often lean bass is mistaken for accurate bass. Don't get me wrong the R3 IS accurate in my estimation, it is also a bit lean before the subs come into play like most bookshelf speakers in this space.
I cross at 68hz and slightly overlap the subs and monitors (this essentially allows for 4 solid bass sources from the 60's to about 120 - filling in much of the nulls from room modes and SBIR), then I use DSP to tame peaks and time align.
The distortion is extremely low and when HP'd like I am doing I'd estimate a complete zero factor.
@Steve Dallas posted his in room energy captures and mine are very similar, the R3 has a very smooth upper energy and a very interesting sound.
Pardon the prose, they are simultaneously dry/clean and neatly rich - like very well made plain dark chocolate. As pristine as they are unfortunately they are never quite fully fun when I rock out. My GF described them as "reverse osmosis water". That said I have zero problem with saying they are indeed 5/5 star speakers and easy to see why they gather high praises (and yet not quite my personal choice @least in my sighted and clumsy fun testing).
I really want to hear the Reference 1 Meta.