From my point of view, HRTEM can do these jobs, but you have to build up model and do simulation of HRTEM image to compare the experimental one, ,, HAADF/STEM maybe more straightforward, simulation maybe not required but always helpful
Why not you read the famous books on TEM? Example, Fultz and Howe, Carter and Williams, etc...If they are not available in your library, even then you may download the books from the internet, although this is a violation of copyright of respective publisher.
Individual vacancies/interstitials might not be possible, but definitely, dislocations, grain boundaries and vacancy clusters can be observed using any of the above mentioned techniques.
Perhaps the most controversial point is about individual vacancy, I remember once I read a paper in Appl. Phys. Lett., there was a report on HRTEM study about individual vacancy in GaN plan-view specimen, I am not sure if my memory is correct, you could search this literature.
I strongly agree with Dear Oğurtani. Please also see our experimental papers for FENICU,FENIMN and FENIMNSICR etc. They will provide you a pre-insight.
Thanks for posting your question, which has initiated such a stimulating discussion. I have been looking for a answer myself. Recently, I came across an online presentation which sort of points in the direction of your question. The file is attached.
Dear Abbas almost halve a century passed by since the discovery of dislocation network in metal foil using TEM/ which was took place in Cambridge Cavendish laboratory, UK by Professor Sir Hirch and his coworkers. I have seen photographs of point defects and their clusters taken by HRTEM in Google platform. But could not copied and pasted them in this RC domain, Regards and Happy New Year!.