I tested all the adapters at three locations in my home using this procedure. Pairs of the same make/model adapters that came in each kit were tested with separate up and downlink tests at each of three test locations. Simultaneous up/downlink and four-stream tests were also run at Location A. All results were entered into the Powerline Performance Charts.
Since past experience has shown that powerline throughput varies with distance, I used the Performance vs. Location view to compare performance. Both adapters are shown, plus two "AV 500" adapters for comparison, the ZyXEL PLA4205 and Belkin Powerline AV 500. (Linksys didn't make an AV 500 product.)
The Downlink Throughput plot below shows both the AV2 products start out with almost a 2-to-1 throughput advantage over the AV 500 products. The Linksys measured a bit higher at 240 Mbps vs. the ZyXEL's 222 Mbps. Remember, "Location A" represents performance with both adapters plugged into the same outlet strip; hardly a real-world use scenario.
Downlink throughput comparison
Moving to Location C, which represents a more typical floor-to-floor deployment, we see that throughput has dropped considerably to the 90 - 100 Mbps range for both adapters. This is still better than the AV 500 products that produced throughputs of 61 and 75 Mbps.
The best news for the AV2 adapters is that their downlink throughput seemed to hold up better than the AV 500 products at worst-case Location E. Both AV2 adapters produced mid 70 Mbps throughput vs. 47 Mbps best case for the AV 500 products.
The IxChariot plot below summarizing the downlink runs for the PLE500 shows a throughput jump around 10 seconds into the test for Location A. This is very similar to what I've found testing draft 802.11ac wireless products. Although Location A throughput is relatively stable, there is more variation than seen at the other test locations. But a good deal of this is probably averaging due to not changing the test file size as throughput dropped.
IxChariot downlink summary - Linksys PLE500
Here's the downlink plot for the ZyXEL PLA5205 pair. We again get a similar jump up toward the start of the Location A test. There is also some cyclical variation that we didn't see in the Linksys plot.
IxChariot downlink summary - ZyXEL PLA5205
Turning to uplink, we see a much different performance picture. The Linksys manages only a slight advantage over the AV 500 products at Location A, producing 142 Mbps vs. their 131 Mbps. In contrast, the PLA5205 stays up at 223 Mbps; almost identical to its downlink results.
It's surprising to see very little difference between the AV2 and AV 500 products at Location C. All four results fell into a 10 Mbps band from 75 to 85 Mbps. At least both AV2 products once again out-performed the AV 500 pair, coming in at 73 Mbps vs. 42 - 46 Mbps.
Uplink throughput comparison
The IxChariot plot below summarizing the uplink runs for the PLE500 once again shows the throughput bump-up near the start of the run. But this time, throughput both starts out lower (120 Mbps vs. 150 Mbps for downlink) and rises only to 145 Mbps or so.
IxChariot uplink summary - Linksys PLE500
The ZyXEL PLA5205 uplink plot looks almost the same as downlink.
IxChariot uplink summary - ZyXEL PLA5205
If you don't like my choices for comparison, use the Powerline Charts and run your own.
I've found that both 200 and 500 Mbps flavors of HomePlug AV can produce higher total bandwidth from multiple streams that you can get from a single stream. AV2, however, seems to provide only a moderate throughput gain from multiple streams. Here's the simultaneous uplink / downlink plot for the Linksys PLE500. The 267 Mbps total is only moderately higher than the best-case 240 Mbps Location A downlink result. The uplink / downlink throughput imbalance is also reflected in the simultaneous test.
Simultaneous up/downlink test - Linksys PLE500
The same test for the ZyXEL had about the same total throughput as the Linksys at 262 Mbps, with a similar moderate gain over its best-base 223 Mbps unidirectional results. In contrast with the Linksys, the ZyXEL has a much more equal distribution between up and downlink.
Simultaneous up/downlink test - ZyXEL PLA5205
It's nice that we finally have an "official" 500 Mbps powerline standard. But since all the "AV 500" adapters used the same Qualcomm Atheros AR1500/AR7400 chipset, AV 500 was a de facto standard anyway.
So are a pair of AV2 500 Mbps (or 600 Mbps as ZyXEL claims) adapters at $120 worth twice the cost of an AV 500 pair? I can only say...maybe. It is true that these two sets of AV2 adapters are capable of almost twice the maximum throughput of most of the AV 500 adapters we tested almost two years ago. But that throughput is obtained only with a same-outlet case that doesn't reflect real-world use.
When the adapters are compared under more typical conditions (our Location C case), both the Linksys and ZyXEL AV2 adapters would be indistinguishable from AV 500 adapters, albeit producing around 80 Mbps of usable throughput. The only real-world advantage our testing has revealed is a significant (almost 2X) throughput advantage for widely-separated adapters (our Location E case). The ~75 Mbps throughput we got with the AV2 adapters might make the difference for uncompressed Blu-ray streaming over a pair of AV 500 adapters supporting only around 40 Mbps.
In the end, my advice is similar to that offered at the end of the AV 500 roundup. First try a pair of AV 500 adapters and move up to either of these AV2 adapters if the AV 500 adapters don't produce the throughput that you want. The additional signal processing magic and techniques in AV2 just might eke out the performance that you need.