I don't think speed is a big factor in the H-20 bomber design.
I think survivability with VLO is the by far the most important design goal.
Keeping costs of production reasonable is definitely be in the design teams' mind, with the B-2 being a prime example of cost overrun.
Making it high speed as well would make it prohibitively expensive, and would negatively impact fleet numbers.
survivability is a function of munitions range, speed and detection range.
Example: dropping unpowered munitions.
Let's say that a bomber is 100% RF stealth - never is but let's just say it is - hits an target with drop bombs. What next? Well, the targeted side will scramble IRST equipped fighters from wherever you could not bomb, like a carrier or another part of the air base.
A subsonic plane flies about 900 kph, about 15 km per minute, let's say. A scrambled fighter going full burn at Mach 2, let's call it 2400 kph for nice round numbers. 40 km per minute. They each can use IRST to see 50 km in each direction, so a 100 km bubble around them.
Within 10 minutes, a subsonic plane can be in a radius of 150 km around the target just bombed. A fighter will cover one axis of the 150 km search area within 4 minutes. They'll be able to use IRST to search half area within 2 minutes. You need only 2 planes on perpendicular axes to find search the entire 150 km radius circle with IRST. Let's say they send 3 fighters. Subsonic plane is not getting away.
There's 3 ways to make a bomber more survivable.
1. use cruise missiles instead of drop munitions (this requires high payload and large internal bay dimensions)
2. make it faster (this requires better or more engines)
3. further reduce signatures (this comes automatically with carrying munitions internally and that gives a huge benefit by itself, but to go further, you have to use less aerodynamically optimal shaping and RAM)
Make no mistake, a Sino Tu-160 White Swan/B-1B will be an absolute White Elephant in today's threat environment.
I would argue that highly visible platforms like the B-1B and Tu-160 have been obsolete since at least the early 1980s. Remember, the U.S Air Force discarded the concept of high altitude high speed bomber such as the XB70 by the early 1960s, when the Soviets began to deploy larger number of long range high performance AAMs.
The follow up B-1 program did not fair any better. It was envisioned to combine the speed of a B-58 Hustler and the payload of a B-52 in low altitude penetrating strikes, but by the early 1970s the Soviet had developed a effective look-down shoot-down radar and rendered such tactics obsolete. Only a combination of U.S domestic politics gamesmanship (Between Carter and Regan) and concerns over ATB Bomber (B-2) program delay saved the B1-B from total cancellation. A relatively low number of 100 was procured and all production ceased by 1988.
Had the Cold War continued apace, the U.S would have acquire far more than mere 21 B-2 Spirit airframes. The original plan called for a purchase of 132+ airframes and the number was slashed only after the dissolution of the Soviet Union. The eye-watering cost of B-2 was a direct result of the dreaded death spiral of program cost rise/procurement cut.
Fast forward to the present day, the trend is clear. All aspects of air power is gravitating toward VLO. Hell, I would not even be surprised if tanker, EW and AEW&C aircraft all goes low observable in the next few decades as well. (The cost would of course be enormous, but one can argue it is already heading that way...)
A B-1B launching LRASMs or a Tu-160 launching Kh-55s is pretty survivable because it has long ranged munitions and has enough speed and range to never be within the range of enemy fighters. the context of a 'low altitude penetrating bomber' is to drop unpowered bombs when air launched cruise missiles weren't very good yet. Tu-160 was never supposed to be a low altitude bomber, it was always a missile truck, and B-1B quickly converted into the missile truck role.
B-1B isn't unstealthy either, it has enough stealth capability (such as the ability to carry all munitions internally and serpentine intakes) that give it a
compared to ~50 m2 for the B-52 and ~25 m2 for a F-15. The biggest contributor to RCS by far is external munitions which act as
Tu-160 also can carry everything internally.
In general for given frequency, range R ~ (RCS)^(1/4) (see pg14). For low hanging fruit, going from say 100 m2 to 1 m2, your detection range decreases 3x. This is not that bad to achieve: a conventional layout with internal munitions, serpentine intakes and some RAM coating gets you close. Going from 1m2 to 0.01 m2 is a further 3x decrease in detection distance but now that requires some extreme modifications to a conventional aerodynamic airframe and imposes constraints on say, speed, fuel efficiency, range, payload weight and dimensions.
Basically it's more complicated than any 1 factor. Realistic requirements and understanding how your plane actually will fit into the overall air doctrine is the most important part. You have to balance cost, reaction speed, payload, range and detection distance.