I think it's all about distance.
Let's make the reasonable assumptions that: (1) LICIAcube uses also X-band to beam data back to Earth; (2) the available transmit power is the same as that on MarCO; and (3) the receive antenna on Earth is the same.
MarCO has to close the link-budget at ~300 millions Km from Earth, while LICIAcube needs to do so at only 11 millions Km (DART impact). Space loss being proportional to the square of distance, that's an advantage of 29dB for LICIAcube.
On the photo of MarCO (given in OP's question), we can see a low-gain 2x2 patch array antenna (same orientation as the high-gain reflectarray, but "sticked" on the front side of the spacecraft body). Such antenna has a gain of 12+ dBi, as can be read on Enduro. Compare this to the 29dBi gain of the reflectarray (see Hodges [1]).
Hence the net link-budget gain for LICIAcube vs MarCO is 12dB (if it uses the Enduro's 2x2 patch array antenna, as I suspect it does, and if everything else are kept identical).
The target bitrate achieved by MarCO is 8 kbps (see Hodges [1]). So, logically LICIAcube can beam down to Earth approximately 16 times MarCO bitrate, i.e. 128kbps.
A paper[2] authored by Simonetti: LICIACube on Dart Mission ... provides corroboratively the following information:
The [communication] subsystem includes an X-Band transponder that manages downlink and uplink communications and is connected to four X-Band patch antennas.
Hence, the X-band is confirmed. Further,
The set of antennas consists of:
- The main pair composed by a transmitting antenna, with 22 dBi gain, and a receiving antenna, with 6 dBi gain, both placed in the solar side structural
panel;
- The secondary pair composed by a transmitting antenna, with 12 dBi gain, and a receiving antenna, with 6 dBi gain, both placed at the opposite side of the main pair.
Not clear yet why they need so many antennas and why are these antennas positioned on opposite sides of the box. But if they intend to use the 22dBi antenna then the bitrate they can achieve will be much higher than 128Kbps, although not 10 times more. Because higher gain induces more stringent pointing accuracy. Note that for MarCO, a 3 dB pointing margin is taken (see Hodges [1]).
The paper does not give the storage capacity, but it mentions
the Argotec HAWK platform (the same used by MarCO by-the-way). According to Argotec data HAWK-6 can have up to 32GB of storage. At 128Kbps download (using the 12dBi antenna as worst scenario), this would require 2 Millions seconds (~560 hours) of connection with a Deep Space Antenna on Earth. From this we can deduce that, at a rate of 4-hour connection per day (rough estimate), we need 140 days to beam back 32GB of data.
1 Hodges et.al. A Deployable High-Gain Antenna Bound to Mars ...
2 Simonetti et.al. LICIACube on DART Mission...