Evidence indicates that the impact angle was about 30 degrees above the horizon, headed east-southeast:

1) Only oblique impacts produce side blowouts.

2) The largest tektite strewn fields are made by 30-degree impacts¹, and the associated Australasian strewn field covers 10% of the earth.

3) There is an east-southeast breach in the central peak.  Such breaches are parallel to the meteorite's trajectory.⁴

4) "In oblique impacts the strength of the shock is asymmetric, with the strongest shock in the downrange direction."³  The downrange shock wave produced the Tonga and Mariana Trenches.

5) The central peak is fairly centered in the crater.  The position of a central peak is not affected by the impact angle.²

6) As discussed elsewhere on this website, the South Pacific Superswell is likely an antipodal effect of the impact, offset due to the obliquity of the impact.

1.  Artemieva, N. 2008. Tektites: Model Versus Reality. 39th Lunar and Planetary Science Conference, p. 1651.
2.  Ekholm, Andreas G., H. Jay Melosh. February 15, 2001. Crater features diagnostic of oblique impacts: The size and position of the central peak. Geophysical Research Letters, Vol. 28, No. 4, pp. 623-626.
3.  Pierazzo, E., H. J. Melosh. 2000. Melt Production in Oblique Impacts. Icarus, Vol. 145, pp. 252-261.
4.  Schultz, P.H., R.R. Anderson. 1996. Asymmetry of the Manson impact structure: Evidence for impact angle and direction, in The Manson Impact Structure, Iowa: Anatomy of an Impact Crater. C. Koeberl and R.R. Anderson (eds). Geological Society of America Special Paper 302, pp. 397-417.

Putting the scattered pieces of Earth's crust back together reveals the hole where the giant meteorite vaporized and excavated the continental crust, including the side blowouts.

The remnants of the crater walls can still be seen as rounded cuts on the coasts of East Africa, northern Madagascar, and southern Australia

The impact was onto continental crust.  But as the land separated, the ocean rushed in and obliterated the crater rim.  This is typical for craters in the ocean.
 Dypvik, Henning, Lubomir F. Jansa. 2003. Sedimentary signatures and processes during marine bolide impacts: a review. Sedimentary Geology, Vol. 161, pp. 309-337.

Yet some effects are evident on the west side.  The coasts of Kenya and Tanzania show severe faulting and diapirism (cracks into which magma seeped).  Numerous normal (pull-apart) faults are downthrown to the east.  The continental shelf is quite narrow (25 to 50 km), with a steep continental slope.¹  Such partial collapse is typical of impacts on the continental margin.²

1.  Coffin, M.F., P.D. Rabinowitz. 1988. Evolution of the conjugate East-African-Madagascan margins and the Western Somali Basin. Geological Society of America Special Paper 226.

2.  Dypvik, Henning, Lubomir F. Jansa. 2003. Sedimentary signatures and processes during marine bolide impacts: a review. Sedimentary Geology, Vol. 161, pp. 309-337.

The West Somali Basin basalt floor has scarcely been sampled.  Drilling was attempted at two sites (240 and 241) in 1972 as part of the Deep Sea Drilling Project.  The drill bit broke after penetrating just 1.2 meters into basalt at site 240, and never reached basalt at site 241.  No one ever tried again.