If it was apples to apples, then show us the full spec of memory, they have a dedicated page to all systems they used, why not put memory specs also?
How exactly would that tell you anything that you don't already know now? They already have the memory specs up on that page. LPDDR4 clocks and timings aren't static like DDR4, they constantly fluctuate with using the laptop, and again, neither of those directly comparable between LPDDR4 and DDR4. If it's an AIDA benchmark run you want Intel to do (which by itself can be laughably bad at representing real-world bandwidth), that would neither seek to further their presentation or our understanding of the product in any way.
Renoir also purportedly supports LPDDR4-4266; the problem is that most Renoir chips end up in thicker laptops with SO-DIMM slots, so no OEMs bother to put in the work to design a LPDDR4-based Renoir laptop because the other 10 SKUs all use soldered or socketed DDR4. You can't test what doesn't exist.
And even had there been a LPDDR4 Renoir ultrabook, I highly doubt the Vega iGPU would see any gains compared to dual-channel DDR4-3200. Same goes for Tiger Lake, there's no reason why LPDDR4-4266 makes it magically faster. LPDDR4 is for power efficiency, not performance. Renoir already achieves fantastic battery life improvements with DDR4.
IIRC, LPDDR4 literally has half the bus width of DDR4.
The point of concern you're looking for here is the continued use of high turbo frequencies that draw power way beyond TDP and PL2. PL1 and PL2 are in the hands of laptop manufacturers, and there's often nothing that can be done if the OEM decides to set PL2 at strictly TDP, which absolutely hobbles your long-term turbo performance. That's something we'll only know from laptop reviewers once the products hit the market.
