True. For comparing, advanced historic research must be provided.

Inventors were living like hermits without much collaboration from third parties whereas trusted acquintances were rare, leading to solo projects. So, 'people engage' is misleading as it implies collaboration like you can find in R&D departments nowadays. Instead, it was much more DIY back then which one can partly (not fully) relate to lack of patent system.

Inventions were stolen and spied upon, inventors made mechanisms to protect their invention otherwise (obfuscation in written specifications, for example) which meant some inventors took their invention with them to their grave. Quite funny in the light of Egyptian pyramids, isn't it? Now, this practice does not mix well with 'constant process of incremental improvements'. How is it an 'incremental improvement' when an invention isn't understood therefore not practically applied? When its hidden, burried? The work was done by the inventor already, yet now someone has to do that work again. Again, lack of ability to write should be taken into consideration as well.

If you think the above are 'rare exceptional cases' you're wrong. It was common practice. Heck, tons of artifacts are still not understood, technologies still have to be reverse engineered, and nobody understands the 7 world wonders. It even happens nowadays in copyright related cases. Countless musicians have lost original artwork due to harddrive failures and Windows reformats, lost or gave away their DATs, CD is not readable anymore, or their unreleased work is given to acquintance who gave it to one or more third parties without permission. You can also see the problem in e.g. Lex Nokia. The problem is an inherent result of data, distribution, and networking which all have become incredibly cheap in past century (and more recent even cheaper ofcourse).