I got Vega 8 benchmarks mixed up when I was comparing Xe to Vega 11. The whole thing is such a mess with Intel's product naming, that I'm constantly getting things confused, so a lot of other benchmark data I saw is probably wrong too when comparing to older hardware. I'm trying to remember the articles I had seen with some benchmark data, but with the laptops plugged in and running full power, the Xe was something like up to 40% faster than Ryzen with Vega 8 on some games and only a bit faster or slower on others.
Looking over other resources showing graphics performance, it may be reasonable to infer that you will get the same or similar benchmark percentages on the lower tier mobile CPUs with Vega and Xe graphics. Honestly though, if you are going to spend extra for an i7u or Ryzen 7u for the iGPU, you may be better off getting a laptop with a lower tier CPU paired with dedicated Geforce MX350 or MX450 for 2020 laptops.
Also, I don't think we will see a GT 3030 anytime soon. If we do get a 3030, it could possibly be based on the mobile MX350 or a cut down MX450 covering the middle ground between the two. I mainly think this is more likely, because they would fit the low power usage of GT 1030 and be of similar TDP. The MX350 is based on the GTX 1050 and the MX450 is based on the GTX 1650. I doubt Nvidia would release the full MX450 as a GT 1030 replacement, because it would cut into sales of the GTX 1650, being only a bit slower, but using much less power. The GTX 1650 is currently the fastest GPU you can buy that is available in low profile and not requiring a PCIE power cable like the GTX 1650 Super.