I don't know if it's fair to say D3T didn't have a lot of time. Development started in 2015, according to the PC Gamer article. Is three years a fair amount of time for a massive port, albeit a massive one? I honestly don't know the answer.
It does seem like an extraordinary long amount of time.
On the other hand, the patches so far have demonstrated that delaying just a couple of months would have resulted in a dramatically better shipping product.
It's weird.
What do we know, though?
1. The source code provided to d3t was incomplete.
2. d3t were not provided with Dreamcast's libraries*.
3. d3t resorted to community tools and original Dreamcast/X-Box discs to obtain and organise certain assets (such as the speech and audio), implying that these were not simply handed to d3t.
4. Up until very recently (most likely post-launch), d3t were not offered consultation with any original Shenmue or Dreamcast staff.
We also know that the entire project has been spearheaded by Sega Europe, not Sega Japan, and that the latter chose a later release date than the former.
Everyone's free to make their own assumptions, but personally I don't think it's reasonable to blame d3t entirely, Sega Europe entirely, or Sega Japan entirely. I think a combination of the three, together with the fact that
this was a difficult project to begin with (Dreamcast ports seem to be difficult in general) that led to the situation we have now.
*the full meaning and implication for this is unclear to me. I take it to mean that d3t had to reverse-engineer, rather than translate, API calls within the code. But I could well be wrong. One thing to keep in mind is that they clearly have had problems emulating the Dreamcast's sound processing chip. They appear to have basically "guessed" their way into things sounding sort of right, most of the time. I wonder if some official documentation for the Dreamcast's hardware would have helped.