mt76: Low tx throughput/High tx retries in STA mode

Has anyone else run into issues with tx throughput via mt76 when it is in STA (WDS/4addr on) mode? (Although I ran into a similar throughput issue when trying to run it as a 802.11s mesh point as well)

Running a mt7612 on a Xiaomi Miwifi Mini, on an OpenWRT snapshot (therefore up to commit d3a589586d1b1d9e1422e31f642a500ff0195f20)

[   12.166396] mt76x2e 0000:01:00.0: card - bus=0x1, slot = 0x0 irq=4
[   12.173039] mt76x2e 0000:01:00.0: ASIC revision: 76120044
[   12.426850] mt76x2e 0000:01:00.0: ROM patch build: 20141115060606a
[   12.439723] mt76x2e 0000:01:00.0: Firmware Version: 0.0.00
[   12.445406] mt76x2e 0000:01:00.0: Build: 1
[   12.449587] mt76x2e 0000:01:00.0: Build Time: 201507311614____
[   12.464409] mt76x2e 0000:01:00.0: Firmware running!

From the mt76 station to the AP is as follows:

[ ID] Interval           Transfer     Bitrate         Retr  Cwnd
[  5]   0.00-1.00   sec   173 KBytes  1.41 Mbits/sec   53   4.24 KBytes
[  5]   1.00-2.00   sec   136 KBytes  1.11 Mbits/sec   75   4.24 KBytes
[  5]   2.00-3.00   sec   414 KBytes  3.39 Mbits/sec   64   4.24 KBytes
[  5]   3.00-4.00   sec   872 KBytes  7.15 Mbits/sec   58   4.24 KBytes
[  5]   4.00-5.00   sec   503 KBytes  4.12 Mbits/sec   48   4.24 KBytes
[  5]   5.00-6.00   sec   749 KBytes  6.14 Mbits/sec   51   4.24 KBytes
[  5]   6.00-7.00   sec  1.05 MBytes  8.84 Mbits/sec   59   2.83 KBytes
[  5]   7.00-8.00   sec   573 KBytes  4.69 Mbits/sec   56   2.83 KBytes
[  5]   8.00-9.00   sec  1024 KBytes  8.39 Mbits/sec   56   5.66 KBytes
[  5]   9.00-10.00  sec   707 KBytes  5.79 Mbits/sec   51   4.24 KBytes
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-10.00  sec  6.08 MBytes  5.10 Mbits/sec  571             sender
[  5]   0.00-10.04  sec  6.03 MBytes  5.04 Mbits/sec                  receiver

However in the other direction, it appears to be fine (From the AP to the STA)

[ ID] Interval           Transfer     Bitrate         Retr  Cwnd
[  5]   0.00-1.03   sec  29.4 MBytes   238 Mbits/sec    0    426 KBytes
[  5]   1.03-2.01   sec  30.2 MBytes   259 Mbits/sec    0    475 KBytes
[  5]   2.01-3.02   sec  24.7 MBytes   205 Mbits/sec    0    475 KBytes
[  5]   3.02-4.01   sec  27.5 MBytes   233 Mbits/sec    0    475 KBytes
[  5]   4.01-5.02   sec  26.2 MBytes   218 Mbits/sec    0    475 KBytes
[  5]   5.02-6.04   sec  30.2 MBytes   250 Mbits/sec    0    475 KBytes
[  5]   6.04-7.02   sec  28.8 MBytes   245 Mbits/sec    0    475 KBytes
[  5]   7.02-8.03   sec  21.2 MBytes   176 Mbits/sec    0    475 KBytes
[  5]   8.03-9.04   sec  20.0 MBytes   165 Mbits/sec    0    475 KBytes
[  5]   9.04-10.01  sec  29.6 MBytes   257 Mbits/sec    0    475 KBytes
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-10.01  sec   268 MBytes   224 Mbits/sec    0             sender
[  5]   0.00-10.06  sec   267 MBytes   223 Mbits/sec                  receiver

And here's the output of iw dev wlan0 station dump:

inactive time:  10 ms
        rx bytes:       50401221
        rx packets:     78919
        tx bytes:       31363544
        tx packets:     25110
        tx retries:     1123
        tx failed:      1
        beacon loss:    0
        beacon rx:      10873
        rx drop misc:   4125
        signal:         -54 [-62, -54] dBm
        signal avg:     -51 [-58, -51] dBm
        beacon signal avg:      -49 dBm
        tx bitrate:     780.0 MBit/s VHT-MCS 8 80MHz short GI VHT-NSS 2
        tx duration:    19562061 us
        rx bitrate:     780.0 MBit/s VHT-MCS 8 80MHz short GI VHT-NSS 2
        rx duration:    15335478 us
        airtime weight: 256
        expected throughput:    67.16Mbps
        authorized:     yes
        authenticated:  yes
        associated:     yes
        preamble:       long
        WMM/WME:        yes
        MFP:            no
        TDLS peer:      no
        DTIM period:    2
        beacon interval:100
        short slot time:yes
        connected time: 1114 seconds

Start by reproducing the reported mt76 STA-mode behavior on the mt7612 Xiaomi Miwifi Mini with WDS/4addr enabled, using the OpenWrt snapshot details and the provided iperf and iw station dump data. Trace the transmit path and compare both traffic directions; done means identifying the cause of the low throughput and high retries and documenting or correcting it.