The principles of Occam's razor and תפסת מרובה לא תפסת are not equivalent. Occam's razor is a logical principle stating that when comparing equivalently predictive hypotheses, the one that requires the fewest assumptions is most likely to be correct.
The principle of "תפסת" on the other hand, is not a logical principle in the same way. Instead, it is more like the principle judicial minimalism. The purpose of judicial minimalism is to limit the scope of rulings to particular cases in order to avoid having to make assumptions about more general laws. For example, the US Supreme Court often tries to resolve cases on narrow grounds specific to the case at hand when possible rather than making a ruling on Constitutional grounds that sets a precedent for future cases. In our case, if our understanding of a law can come from two different sources, we should understand it to come from whichever source is narrower (i.e. leads to us making the fewest assumptions). This way, in the case that we were wrong about from where the law should be learned, we avoid doing the wrong thing. The analogy is not perfect since in the Talmud we are not discussing judicial rulings, but it is sufficient to explain the difference from Occam's razor.
In your example from Yoma, the discussion is the size of an egg. It is suggested that maybe the egg-size is based on the size of an egg from the extremely large bar-yochani bird. The problem is that if we are incorrect about that and really egg-size is based on the egg of a smaller bird, we are at risk of doing something wrong with food between the size of the two eggs. Instead, we rule that egg-size comes from the smallest kind of egg because if we are wrong, it can only make things less likely that we will make a mistake.
So in summary, Occam's razor states that the hypothesis that requires the fewest assumptions is most likely correct, whereas תפסת מרובה לא תפסת means that we try to limit our halachic rulings in order to avoid making assumptions and consequently possibly doing the wrong thing. It does not attempt to figure out "which hypothesis is correct" but rather "which hypothesis is safer."