The 89C2051 datasheet says Iol is limited to 10mA per port
line and one drive LED directly.
This may be different for 89c51 or other micros.
Also it is not good scare novice person with some
irrevelent assuptions.



From TB62726 datasheet I found following Input specifications:
Input signal voltage level: 3.3-V and 5-V CMOS level (Schmitt
trigger input)
Max clock rise time = 5uS.
Other propogation delays ETC are appx 300nS.

Hazman has pullup resistor of 1Kohm so risetime is
1x10^3 x 30 x 10^-12 = 30 x 10^-9 i.e. 30nS. where 30pf
is assumed of micro + TB pin capacitance.
Therefore this well acceptable even for critical TTL devices.
I use 4.7K as pullup which is many manufacturers are recommending taking in view the economy of current supply and
AC timing.
It is not good to scare Hazman for using port0 if he wants
to. All he has to adhere to basics and most importantly devices
being used.
It is always good policy to limit LED current with serries
resistor rather than rely on other devices. I have found people
making multiplexed LED displays with 2051micros using 33ohms in
5V circuits. In other circumstances this would drive about
3.2/.033 = appx 100mA ! But no! The LEDs draw only 20mA per segment! When I raised the resistor value to 150 ohms same
intensity was seen. One should note that micro was running
cooler then. I am not sure about 89c51 device. I have seen
that peole use rather ULN 2803 for that purpose.
Hope that point is understood.
1. There is no harm in using port 0 with proper pull ups
2. Certain micros are (Including many from Microchip) designed
for direct driving LEDs as their current sinking capacity
is limited to 10~20mA.