The performance from this circuit is very satisfactory. The sheet contains data for typical noise and distortion statistics, I cannot confirm these statistics because I don't have the equipment, but my ears tell me that the sound is good and with sensitive speakers, this amp will show some power. As can be seen, the circuit is quite simple indeed. If you can make your own PCB he circuit is very simple to make and you can design the PCB to accept the odd, staggered pin-outs of some TO220 devices.
Resistors should be at least 1/4W types with 1% tolerance recommended. I used 0.6W 1% metal film resistors and they work well. Capacitors I used were electrolytic cans for C2, C5 and C6. At the time of building, I did not have enough 100uF caps so I used 220uF caps instead, this will not cause trouble, unless you use caps that are smaller than 100uF for C5 and C6.
C1 can be electrolytic, I used a tantalum myself (don't ask why, since they are actually more expensive). Some readers may want to use a polyester capacitor for the input (C1), this would work too but I am not sure whether any benefit will arise from the extra expense. Other capacitors (bypass caps C3 and C4. and C7) are ideally polyester caps. Mylar may be used but I prefer polyester caps and they are not expensive unless you get large values.
The values of R5 and C8 are worked out from the equations in the datasheet, but I used 1.8k ohm for R5 and 220pF for C8 and they work fine. Diodes should be 1N4001 or similar (make sure you solder them in the right way round).
A good heatsink is important and this should be a large size with good thermal conductivity. When you operate the TDA2030 from the (recommended) split rail power supply, you must insulate the device from the heatsink using a mica washer or similar. With single rail supply, this is not needed.