Just a couple of observations (having been involved in a number of patent litigations through trial and appeal): paulr is correct in saying: "Ultimately patents are only as good as the amount of time and money you want to put into enforcing them." One should also keep in mind that litigations involving patents tend to be more expensive than "normal" litigations for reasons that include, but are not limited, to the following: Patent litigations almost always require the participation of expert witnesses (usually at least two per side, one for technical issue and one for damages) who, in my personal experience, can charge up to $900 per hour; every patent case involves "construing" the claims of the patent in suit -- this usually involves what is commonly referred to as a "Markman hearing" which can be quite expensive, sometimes amounting to a sort of "mini-trial" complete with expert testimony and the attendant preparation, expensive demonstrative exhibits and plenty of attorney time; discovery tend to be more voluminous and often include the depositions of more witnesses than in other types of cases -- some of the cases I have been involved with have required prying the lids off of and examining the contents of literally thousands of boxes of documents; the prospect of repeated appeals in the same action is very real -- I have personally been involved in actions that have been to the Court of Appeals for the Federal Circuit (the court that hears all appeals in patent actions) multiple times; finally, patent suits with their attendant complexity of issues often take longer than other suits -- some of the actions can take more than 10 years. All this can be quite expensive -- it is not unreasonable to expect that the cost of a patent action, through verdict (but not appeal) can easily exceed $1M.
That said, there are other reasons for obtaining patents than expecting to litigate them. For example, a portfolio of patents may enhance the value of a company that is being or going to be sold.
Regarding patenting something yourself, keep in mind that if the patent is on something that will be a real money maker it may very well be the subject of litigation (or, if you are licensing it, subject to close scrutiny by potential licensees). The process of obtaining a patent in the US (called patent "prosecution") involves a number of pitfalls for the unwary -- an example: statements made during the prosecution of the patent, or even adjustments of the wording of claims may lead to "prosecution history estoppel" that limits the scope of the claimed invention.
Re discussions of potentially patentable subject matter. Section 102(b) of Title 35 of the US Code sets a statutory bar on patentability of inventions in "public use" or "on sale" more than one year prior to the filing of an application for patent. What constitutes a "public use" and potential exceptions to the bar may be difficult for the non-initaited to discern and are a pitfall that form the nucleus of invalitity contentions in every patent litigation I have been involved in.
Regarding getting priority for an invention -- the issue of what constitutes the patentable subject matter is usually a matter of contention in litigations. The scope of the invention may change before or during the process of prosecution because of the prior art, etc. Sometimes, therefore, what the inventor actually initially believes to be the invention does not coincide with that which is claimed.
One should always seek the advice of competent legal counsel on these matters. This, unfortunately, usually brings added expense.