Oh boy... has this turned into a brand-name and flame war! I am both a tech and long time end user. Here is my two cents and I'm done.
On the original poll question:
1. Audible alert on a weak battery: Yes, always, and repeated every few minutes and on PTT.
2. Bloody murder audible alert on a flat battery: Yes, always, repeated frequently and on PTT.
3. Full shutdown to a brick condition (including RX), OR low TX power and some function: Low TX and some function with obvious user notification.
Opinion: (assume the user was NOT taking a knife to a gunfight)
1. Even a subtle beating over the head is sometimes ignored or unnoticed in the heat of the battle.
2. I would rather be able to hear that nobody answered, hear a "repeat your traffic", or have a 3rd party relay my message than have a brick or hear blissful silence.
3. A brick will not convey an evacuation order, but a RX-only portable will.
4. I would rather only have my lead-in ANI show up with no voice following than nothing at all.
5. I would rather be able to talk simplex 50 feet to another responder that can relay me than not be able to talk at all.
First Proviso: All inspection, maintenance and due diligence was performed prior to the failure.
Second Proviso: ILDH or tactical life and safety critical communications are not requiring use of remote infrastructure due to the additional points of failure.
Third Proviso: Due to an un-diagnosed battery failure there is not enough available battery energy to send communication any further than reduced watts will go. i.e.: You can't squeeze blood from a turnip, VS. arbitrarily reducing power because the battery is low.
On anything other than using plain stupid simplex analog in any tactical or IDLH environment: LODD report contributors back this up, while non-line actors want more sales, un-needed bells and whistles, off-site micro management or CYA distant archiving of all tactical traffic.
The purpose of the exercise is reliable communication inside a hot zone and with Safety / Command / Other nearby. Any introduction of un-necessary complexity adds a potential layer of failure to emergency and mayday traffic. A digital radio faced with interference of the desired signal blissfully does nothing while a stupid analog radio will play something, even if unintelligible. That is the nature of an over-BER digital failure vs an interfering signal on stupid analog. A portable radio user in an IDLH calling less than 300 feet does not need extra layers of failure by associating and requesting talk-group access from a repeater miles away over bad road.
The same goes for command sending an emergency message 300 feet but having critical traffic folded, spindled and mutilated through a distant repeater (with -85db downlink signal in the hot zone) when the uplink signal from his radio covers the hot zone at -60db... and may be the interference wacking out his own message.
On radio users checking their equipment: The Captain, Chief or department tech does not personally check an end user's flashlight, 4-gas meter, fuel tank and BA prior to every call for service. That responsibility lies solely with the end user. The same applies to radios - any brand radio - and the batteries that power them.
Just because you have a tool - any tool - does not mean that it will work until you know it will work. Knowledgeable users check every critical aspect of every tool's operation every time - period - and that is a result of training, familiarity with equipment, and in many cases habit. Verifying that all tools are fully functional before performing any tactical or IDLH operation is the user's responsibility, even if other's have recently done the same check. I refuse to bet my life on you remembering not to forget.
On alerting the user of an issue: BA's have PASS alarms and other indicators. New NFPA apparatus alert to seat belt non-compliance. Saws do not have gas gauges, but if it feels light and is sputtering there's your sign. Flashlight go dim. Nearly all new-er radios (not just those where the model number and price equate) can provide an audible alert that the battery is low, and no user - especially those with critical need - should ignore those alerts.
That the 'old radios didn't do that' or users do not know what the alert means is a failure of the command structure above the user, and a failure of the user to be familiar with all aspects of his tools. An ignored alert is NOT not a failure of the equipment. Cliche': Make something idiot proof and someone will come out with better idiots.
On battery failures leading to PA cutoff: Key word is failures. Verifying that all tools are fully functional before performing any tactical or IDLH operation is the user's responsibility.
All three brands of radios I deal with have perfect batteries that never fail. I also have a bridge for sale.
Regular checking, conditioning and replacement will NEVER catch all battery failures. Battery failures happen without warning. I can't find any 72-month-super-quality vehicle battery that is 100% reliable after two years. All infrastructure backup and standby generator batteries are replaced at three years, no matter what.
A typical 2-way transmission lasts <5 seconds. A radio that sees 7.4 volts when the PTT is pushed has no time to warn the user that a dead cell drops the volts to 6.0 two seconds into transmit. No PA will stay up at that low voltage, and the user has seconds to react, often while focused on more critical things.
Dead cells can only be identified by regular battery evaluation and red flagging a battery that failed in use. Battery failures are almost always from batteries past their age or cycle lifespan and never checked for sustained performance. Another reason every radio should have an immediately available hot battery that is swapped prior to IDLH operations, with more batteries on standby for backup.
Intense operation batteries should be swapped at every rehab.
On brands: You generally get what you pay for, but paying four times more does not guarantee you get four times better. Not all models from any one manufacturer perform the same. There are thousands of 15 Y.O. Icom F11's still in service, and boxes of dead F3's. The Spectra Easter Egg capacitor hatches every day. VX-2100's that are perfect in wide are stone deaf in narrow. Every manufacturer has radios they would love to forget they ever built. EVERY manufacturer.
Not all dealers program radios to work properly or to full capacity. Not all agencies spec or program radios correctly... and politics kill responders. I have been to customers with mid-tier radios - other than my primary brand - who complained their equipment was junk. 3 out of 5 times only the batteries were junk or they were simply programmed wrong (Note: in two recent cases radios had State-Mandated programming that guaranteed failure of the State-Tech's NON-preferred brand.)
A department with a 65 call/year volume (two thirds of them BS or 'frequent flyer' calls) won't qualify for big money grants. An agency with an annual $20K budget can't spend the wad on radios, and is hard pressed to pony up their 5% match for a $100K grant for new equipment. That is an unavoidable fact of life. Impres(r) is a good idea, but beyond many budgets, and multiple model radios in the same environment lead to end-user confusion and extra parts to support them all.
Some dealers put in frequency, CTCSS and call it good. Better dealers take the time to go through every screen and set radios up right as verified by first hand experience before releasing product to the customer, then check real world operation over time . The best dealers will not sell the wrong radio or radios not set up properly... to anyone at at any price.