The issue is the the halving may cause an large abrupt drop in the total hash rate, and blocks would take longer until the difficulty adjusts.
This year, the difficulty is nominally scheduled to adjust a week after the halving. Suppose that the hash rate drops by half (not likely, but not impossible). In that case, blocks will each take 20 minutes and the difficulty won't change for two weeks. Cutting Bitcoin's transaction rate by half for two weeks could have a major impact.
It is nearly the same as what we saw in 2017 when bitcoin transactions got stucked even people tried to boost their fees to higher and higher. That's terrible in that period but such things often occur whenever bitcoin price skyrockets to the peaks. People turn to more actively move their coins and accept higher fees while more miners join. After all, when price drops back and newbie miners leave, we see some kinds of effects from less total hashrate and a 14 days waiting period to see difficulty adjusts. This year, the difficulty is nominally scheduled to adjust a week after the halving. Suppose that the hash rate drops by half (not likely, but not impossible). In that case, blocks will each take 20 minutes and the difficulty won't change for two weeks. Cutting Bitcoin's transaction rate by half for two weeks could have a major impact.
